Starting Xolotl (2.4.0-83-e84edee7) Thu Nov 30 09:48:13 2023 NetworkHandler: Loaded network of 312 DOF with: Helium Vacancy Interstitial MaterialHandler: The selected material is: W111 with the following processes: advec attenuation diff modifiedTM movingSurface reaction ; a custom fit flux handler is used reading: tridyn_benchmark_PSI_9.dat TemperatureHandler: Using the time profile defined in: temp_benchmark_PSI_9.dat SolverHandler: 1D simulation with surface BC: free surface and bulk BC: free surface, initial concentration for Id: 0 of: 1e-18 nm-3, grid (nm): 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.7999999999999999 0.8999999999999999 0.9999999999999999 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 2.100000000000001 2.200000000000001 2.300000000000001 2.400000000000001 2.500000000000001 2.750000000000001 3.000000000000001 3.250000000000001 3.500000000000001 3.750000000000001 4.000000000000002 4.250000000000002 4.500000000000002 4.750000000000002 5.000000000000002 0 TS dt 1e-10 time 0. Time: 0 Helium content = 0 Vacancy content = 4.750000000000002e-18 Interstitial content = 0 Fluence = 0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 1 TS dt 1e-09 time 1e-10 Time: 1e-10 Helium content = 2.424952199503062e-06 Vacancy content = 6.71808169858663e-15 Interstitial content = 3.133080177694656e-09 Fluence = 5.4e-06 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 2 TS dt 1e-08 time 1.1e-09 Time: 1.1e-09 Helium content = 2.296916600451812e-05 Vacancy content = 2.31356670859338e-12 Interstitial content = 3.19440449167048e-09 Fluence = 5.940000000000001e-05 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 3 TS dt 6.80732e-08 time 1.11e-08 Time: 1.11e-08 Helium content = 0.0001055007230322544 Vacancy content = 3.823043552405469e-09 Interstitial content = 3.218807071399915e-09 Fluence = 0.0005994 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 4 TS dt 5.05362e-07 time 7.91732e-08 Time: 7.917317101260537e-08 Helium content = 0.0001189610844555452 Vacancy content = 3.906703873501566e-07 Interstitial content = 3.444351103043243e-09 Fluence = 0.00427535123468069 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 5 TS dt 5.05362e-06 time 5.84535e-07 Time: 5.845350088239711e-07 Helium content = 0.0001258488093923175 Vacancy content = 5.658565060233405e-06 Interstitial content = 3.489696739356304e-09 Fluence = 0.03156489047649444 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 6 TS dt 1e-05 time 5.63815e-06 Time: 5.638153386937629e-06 Helium content = 0.0002010574732121331 Vacancy content = 5.316793387390591e-05 Interstitial content = 3.364559413993797e-09 Fluence = 0.304460282894632 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 7 TS dt 1e-05 time 1.56382e-05 Time: 1.563815338693763e-05 Helium content = 0.0003315813442163533 Vacancy content = 0.0001231337003572065 Interstitial content = 3.19428926503291e-09 Fluence = 0.8444602828946322 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 8 TS dt 1e-05 time 2.56382e-05 Time: 2.563815338693763e-05 Helium content = 0.0004476286196065412 Vacancy content = 0.0001729974611806802 Interstitial content = 3.07534661688012e-09 Fluence = 1.384460282894632 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 9 TS dt 1e-05 time 3.56382e-05 Time: 3.563815338693763e-05 Helium content = 0.0005560483217551426 Vacancy content = 0.0002106012850672489 Interstitial content = 2.986545711612676e-09 Fluence = 1.924460282894632 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 10 TS dt 1e-05 time 4.56382e-05 Time: 4.563815338693763e-05 Helium content = 0.0006606877895992233 Vacancy content = 0.0002402069186172472 Interstitial content = 2.917129573419367e-09 Fluence = 2.464460282894632 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 11 TS dt 1e-05 time 5.56382e-05 Time: 5.563815338693763e-05 Helium content = 0.0007637911332658899 Vacancy content = 0.0002643132393436003 Interstitial content = 2.861037068682926e-09 Fluence = 3.004460282894633 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 12 TS dt 1e-05 time 6.56382e-05 Time: 6.563815338693763e-05 Helium content = 0.0008667085407284311 Vacancy content = 0.0002844785068171919 Interstitial content = 2.814578315379641e-09 Fluence = 3.544460282894633 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 13 TS dt 1e-05 time 7.56382e-05 Time: 7.563815338693763e-05 Helium content = 0.0009702704188514846 Vacancy content = 0.000301727522122802 Interstitial content = 2.775369688878707e-09 Fluence = 4.084460282894632 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 14 TS dt 1e-05 time 8.56382e-05 Time: 8.563815338693763e-05 Helium content = 0.001074993517662594 Vacancy content = 0.0003167689818082796 Interstitial content = 2.741800128508006e-09 Fluence = 4.624460282894632 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 15 TS dt 1e-05 time 9.56382e-05 Time: 9.563815338693763e-05 Helium content = 0.001181198912218485 Vacancy content = 0.0003301189984602513 Interstitial content = 2.712742195940176e-09 Fluence = 5.164460282894632 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 16 TS dt 1e-05 time 0.000105638 Time: 0.0001056381533869376 Helium content = 0.001289081649536771 Vacancy content = 0.0003421747902000013 Interstitial content = 2.687384476721404e-09 Fluence = 5.704460282894632 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 17 TS dt 1e-05 time 0.000115638 Time: 0.0001156381533869376 Helium content = 0.001398752812275666 Vacancy content = 0.0003532599222239556 Interstitial content = 2.66512810034976e-09 Fluence = 6.244460282894632 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 18 TS dt 1e-05 time 0.000125638 Time: 0.0001256381533869376 Helium content = 0.001510265353240789 Vacancy content = 0.0003636520972540441 Interstitial content = 2.645519412600572e-09 Fluence = 6.784460282894633 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 19 TS dt 1e-05 time 0.000135638 Time: 0.0001356381533869376 Helium content = 0.001623630275977358 Vacancy content = 0.0003735995944935821 Interstitial content = 2.628204479217124e-09 Fluence = 7.324460282894633 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 20 TS dt 1e-05 time 0.000145638 Time: 0.0001456381533869376 Helium content = 0.001738827119278414 Vacancy content = 0.0003833301014514786 Interstitial content = 2.612897862716965e-09 Fluence = 7.864460282894633 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 21 TS dt 1e-05 time 0.000155638 Time: 0.0001556381533869376 Helium content = 0.001855811114887512 Vacancy content = 0.0003930545495852893 Interstitial content = 2.599361401953524e-09 Fluence = 8.404460282894632 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 22 TS dt 1e-05 time 0.000165638 Time: 0.0001656381533869376 Helium content = 0.00197451801008536 Vacancy content = 0.0004029679290481984 Interstitial content = 2.587390239560163e-09 Fluence = 8.944460282894632 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 23 TS dt 1e-05 time 0.000175638 Time: 0.0001756381533869376 Helium content = 0.002094865846606066 Vacancy content = 0.0004132485666126602 Interstitial content = 2.576803945033726e-09 Fluence = 9.484460282894631 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 24 TS dt 1e-05 time 0.000185638 Time: 0.0001856381533869376 Helium content = 0.00221675055459619 Vacancy content = 0.0004240568281110305 Interstitial content = 2.567440755652846e-09 Fluence = 10.02446028289463 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 25 TS dt 1e-05 time 0.000195638 Time: 0.0001956381533869376 Helium content = 0.002339757010735212 Vacancy content = 0.0004353587957753541 Interstitial content = 2.560356349853614e-09 Fluence = 10.56446028289463 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 26 TS dt 1e-05 time 0.000205638 Time: 0.0002056381533869376 Helium content = 0.002465509064965822 Vacancy content = 0.0004484855911997163 Interstitial content = 2.559284490854252e-09 Fluence = 11.10446028289463 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 27 TS dt 1e-05 time 0.000215638 Time: 0.0002156381533869376 Helium content = 0.002593525609007556 Vacancy content = 0.0004632835379792119 Interstitial content = 2.563163403775028e-09 Fluence = 11.64446028289463 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 28 TS dt 1e-05 time 0.000225638 Time: 0.0002256381533869376 Helium content = 0.002724034355062431 Vacancy content = 0.0004799838561594826 Interstitial content = 2.567424698324675e-09 Fluence = 12.18446028289463 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 29 TS dt 1e-05 time 0.000235638 Time: 0.0002356381533869376 Helium content = 0.002857459707245525 Vacancy content = 0.0004988266047559642 Interstitial content = 2.572168368229206e-09 Fluence = 12.72446028289463 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 30 TS dt 1e-05 time 0.000245638 Time: 0.0002456381533869377 Helium content = 0.002994370187094124 Vacancy content = 0.0005201303348919841 Interstitial content = 2.577560955910943e-09 Fluence = 13.26446028289463 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 31 TS dt 1e-05 time 0.000255638 Time: 0.0002556381533869377 Helium content = 0.0031355364822675 Vacancy content = 0.000544334096574835 Interstitial content = 2.583857821748516e-09 Fluence = 13.80446028289463 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 32 TS dt 1e-05 time 0.000265638 Time: 0.0002656381533869377 Helium content = 0.003282012986122711 Vacancy content = 0.0005720544117433223 Interstitial content = 2.591434559232261e-09 Fluence = 14.34446028289463 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 33 TS dt 1e-05 time 0.000275638 Time: 0.0002756381533869377 Helium content = 0.003435251906943356 Vacancy content = 0.0006041631318809235 Interstitial content = 2.600830756904739e-09 Fluence = 14.88446028289463 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 34 TS dt 1e-05 time 0.000285638 Time: 0.0002856381533869378 Helium content = 0.003597262513369005 Vacancy content = 0.0006418942534852459 Interstitial content = 2.612810634859524e-09 Fluence = 15.42446028289463 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 35 TS dt 1e-05 time 0.000295638 Time: 0.0002956381533869378 Helium content = 0.003770232089351205 Vacancy content = 0.0006864766983993774 Interstitial content = 2.628167104194366e-09 Fluence = 15.96446028289463 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 36 TS dt 1e-05 time 0.000305638 Time: 0.0003056381533869378 Helium content = 0.003958238589176528 Vacancy content = 0.0007404855168383525 Interstitial content = 2.654075089372328e-09 Fluence = 16.50446028289463 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 37 TS dt 1e-05 time 0.000315638 Time: 0.0003156381533869378 Helium content = 0.004139532180905154 Vacancy content = 0.0007827827581327699 Interstitial content = 2.628571414623529e-09 Fluence = 17.04446028289464 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 38 TS dt 1e-05 time 0.000325638 Time: 0.0003256381533869379 Helium content = 0.004322983495814651 Vacancy content = 0.0008236313750226637 Interstitial content = 2.645245463163054e-09 Fluence = 17.58446028289464 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 39 TS dt 1e-05 time 0.000335638 Time: 0.0003356381533869379 Helium content = 0.00451257705221086 Vacancy content = 0.0008671785596303175 Interstitial content = 2.663027510894321e-09 Fluence = 18.12446028289464 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 40 TS dt 1e-05 time 0.000345638 Time: 0.0003456381533869379 Helium content = 0.00470870580030648 Vacancy content = 0.0009135015211718177 Interstitial content = 2.682059197538296e-09 Fluence = 18.66446028289464 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 41 TS dt 1e-05 time 0.000355638 Time: 0.0003556381533869379 Helium content = 0.004911798713845195 Vacancy content = 0.0009626929139013768 Interstitial content = 2.702531605968983e-09 Fluence = 19.20446028289465 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 42 TS dt 1e-05 time 0.000365638 Time: 0.000365638153386938 Helium content = 0.005122321788583731 Vacancy content = 0.001014864404341224 Interstitial content = 2.724674628689288e-09 Fluence = 19.74446028289465 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 43 TS dt 1e-05 time 0.000375638 Time: 0.000375638153386938 Helium content = 0.005340786761217188 Vacancy content = 0.001070150451130229 Interstitial content = 2.74876559889709e-09 Fluence = 20.28446028289465 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 44 TS dt 1e-05 time 0.000385638 Time: 0.000385638153386938 Helium content = 0.005567765323462502 Vacancy content = 0.001128712631895018 Interstitial content = 2.775141488393701e-09 Fluence = 20.82446028289466 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 45 TS dt 1e-05 time 0.000395638 Time: 0.000395638153386938 Helium content = 0.00580390937317407 Vacancy content = 0.001190745259918844 Interstitial content = 2.804216993218947e-09 Fluence = 21.36446028289466 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 46 TS dt 1e-05 time 0.000405638 Time: 0.0004056381533869381 Helium content = 0.006049980429867046 Vacancy content = 0.001256483679967215 Interstitial content = 2.836511925052035e-09 Fluence = 21.90446028289466 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 47 TS dt 1e-05 time 0.000415638 Time: 0.0004156381533869381 Helium content = 0.006306894494287616 Vacancy content = 0.001326217583032197 Interstitial content = 2.872693213518214e-09 Fluence = 22.44446028289466 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 48 TS dt 1e-05 time 0.000425638 Time: 0.0004256381533869381 Helium content = 0.006575793505168974 Vacancy content = 0.00140031323246831 Interstitial content = 2.913640511641545e-09 Fluence = 22.98446028289467 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 49 TS dt 1e-05 time 0.000435638 Time: 0.0004356381533869382 Helium content = 0.006858163417956545 Vacancy content = 0.001479251371861391 Interstitial content = 2.960551627823275e-09 Fluence = 23.52446028289467 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 50 TS dt 1e-05 time 0.000445638 Time: 0.0004456381533869382 Helium content = 0.00715603606123959 Vacancy content = 0.001563693095010153 Interstitial content = 3.015118615075782e-09 Fluence = 24.06446028289467 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 51 TS dt 1e-05 time 0.000455638 Time: 0.0004556381533869382 Helium content = 0.007472347434810338 Vacancy content = 0.00165459723792622 Interstitial content = 3.079836752663914e-09 Fluence = 24.60446028289467 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 52 TS dt 1e-05 time 0.000465638 Time: 0.0004656381533869382 Helium content = 0.007811603076341338 Vacancy content = 0.001753437069540143 Interstitial content = 3.158580279676169e-09 Fluence = 25.14446028289468 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 53 TS dt 1e-05 time 0.000475638 Time: 0.0004756381533869383 Helium content = 0.008181184798735803 Vacancy content = 0.001862619430467793 Interstitial content = 3.257755641637142e-09 Fluence = 25.68446028289468 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 54 TS dt 7.18092e-06 time 0.000485638 Time: 0.0004856381533869383 Helium content = 0.008594102836158505 Vacancy content = 0.001986345473465493 Interstitial content = 3.388819398835117e-09 Fluence = 26.22446028289468 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 55 TS dt 7.18092e-06 time 0.000492819 Time: 0.0004928190766934691 Helium content = 0.00893047196645343 Vacancy content = 0.002088298687205072 Interstitial content = 3.514456458882748e-09 Fluence = 26.61223014144734 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 KSP Object: 4 MPI processes type: gmres restart=30, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement happy breakdown tolerance 1e-30 maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using PRECONDITIONED norm type for convergence test PC Object: 4 MPI processes type: fieldsplit FieldSplit with MULTIPLICATIVE composition: total splits = 2, blocksize = 313 Solver info for each split is in the following KSP objects: Split number 0 Defined by IS KSP Object: (fieldsplit_0_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_0_) 4 MPI processes type: jacobi type DIAGONAL linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 Split number 1 Defined by IS KSP Object: (fieldsplit_1_) 4 MPI processes type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_) 4 MPI processes type: redundant First (color=0) of 4 PCs follows KSP Object: (fieldsplit_1_redundant_) 1 MPI process type: preonly maximum iterations=10000, initial guess is zero tolerances: relative=1e-05, absolute=1e-50, divergence=10000. left preconditioning using NONE norm type for convergence test PC Object: (fieldsplit_1_redundant_) 1 MPI process type: lu out-of-place factorization tolerance for zero pivot 2.22045e-14 matrix ordering: nd factor fill ratio given 5., needed 3.17245 Factored matrix follows: Mat Object: (fieldsplit_1_redundant_) 1 MPI process type: seqaijkokkos rows=11268, cols=11268 package used to perform factorization: kokkos total: nonzeros=1266817, allocated nonzeros=1266817 not using I-node routines linear system matrix = precond matrix: Mat Object: 1 MPI process type: seqaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node routines linear system matrix = precond matrix: Mat Object: (fieldsplit_1_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268 total: nonzeros=399318, allocated nonzeros=399318 total number of mallocs used during MatSetValues calls=0 not using I-node (on process 0) routines linear system matrix = precond matrix: Mat Object: Mat_0x556cd9a44ae0_0 (fieldsplit_0_) 4 MPI processes type: mpiaijkokkos rows=11268, cols=11268, bs=313 total: nonzeros=399318, allocated nonzeros=0 total number of mallocs used during MatSetValues calls=0 56 TS dt 1e-05 time 0.0005 Time: 0.0005 Helium content = 0.009318381421541287 Vacancy content = 0.002206840625499091 Interstitial content = 3.683255457749294e-09 Fluence = 27.00000000000001 --- Timers: Flux: process_count: 4 min: 2.57282805500001 max: 2.673946429000008 average: 2.625496042750001 stdev: 0.04738531765081894 Partial Derivatives: process_count: 4 min: 2.023499907999995 max: 2.288700306 average: 2.136961279749999 stdev: 0.1090961292067129 monitor1D:checkNeg: process_count: 4 min: 0 max: 0 average: 0 stdev: 0 monitor1D:event: process_count: 4 min: 0.003643629999999999 max: 0.003649098 average: 0.0036460355 stdev: 1.961620439678616e-06 monitor1D:heRet: process_count: 4 min: 0.007963680999999998 max: 0.008958236 average: 0.008215529499999999 stdev: 0.0004288129133203095 monitor1D:init: process_count: 4 min: 1.9974e-05 max: 7.3905e-05 average: 3.35945e-05 stdev: 2.327379309975064e-05 monitor1D:postEvent: process_count: 4 min: 0 max: 0 average: 0 stdev: 0 monitor1D:scatter: process_count: 4 min: 0 max: 0 average: 0 stdev: 0 monitor1D:series: process_count: 4 min: 0 max: 0 average: 0 stdev: 0 monitor1D:startStop: process_count: 4 min: 0 max: 0 average: 0 stdev: 0 monitor1D:tridyn: process_count: 4 min: 0 max: 0 average: 0 stdev: 0 monitor1D:xeRet: process_count: 4 min: 0 max: 0 average: 0 stdev: 0 rhsFunctionTimer: process_count: 4 min: 2.745660812 max: 2.883327671 average: 2.81164295975 stdev: 0.06020129453595313 rhsJacobianTimer: process_count: 4 min: 2.348997845 max: 2.554881655 average: 2.4384714175 stdev: 0.08492036181598517 solveTimer: process_count: 4 min: 31.559086476 max: 31.559115041 average: 31.55910728175 stdev: 1.203010159639632e-05 Counters: Flux: process_count: 4 min: 4448 max: 5004 average: 4726 stdev: 278 Partial Derivatives: process_count: 4 min: 2656 max: 2988 average: 2822 stdev: 166