[petsc-users] PCASM in transposed format

Antoine De Blois antoine.deblois at aero.bombardier.com
Mon Sep 22 13:28:29 CDT 2014


Dear all,

I am using the ASM preconditioner to solve a transposed system through MatSolveTranspose. Strangely, the results I obtain differ in each call. Is there a non-deterministic operations within ASM? If I use it in a non-transposed way, I get correct results... If I use GASM, then the results are always the same, both in transposed and non-transposed formats.

Below is a log of my calls (note that some of them actually diverged).
I can give you my matrix and rhs if you want.

Regards,
Antoine

$ mpirun -np 4 /gpfs/fs1/aero/SOFTWARE/TOOLS/PROGRAMMING/petsc/src/ksp/ksp/examples/tutorials/ex10 -f0 A_and_rhs.bin  -trans -pc_type asm
Number of iterations = 690
Residual norm 0.000617544

$ mpirun -np 4 /gpfs/fs1/aero/SOFTWARE/TOOLS/PROGRAMMING/petsc/src/ksp/ksp/examples/tutorials/ex10 -f0 A_and_rhs.bin  -trans -pc_type asm
Number of iterations = 475
Residual norm 0.000727253

$ mpirun -np 4 /gpfs/fs1/aero/SOFTWARE/TOOLS/PROGRAMMING/petsc/src/ksp/ksp/examples/tutorials/ex10 -f0 A_and_rhs.bin  -trans -pc_type asm
Number of iterations = 10000
Residual norm 1.3866

$ mpirun -np 4 /gpfs/fs1/aero/SOFTWARE/TOOLS/PROGRAMMING/petsc/src/ksp/ksp/examples/tutorials/ex10 -f0 A_and_rhs.bin  -trans -pc_type asm
Number of iterations = 568
Residual norm 0.000684401

$ mpirun -np 4 /gpfs/fs1/aero/SOFTWARE/TOOLS/PROGRAMMING/petsc/src/ksp/ksp/examples/tutorials/ex10 -f0 A_and_rhs.bin  -trans -pc_type asm
Number of iterations = 540
Residual norm 0.000555548

$ mpirun -np 4 /gpfs/fs1/aero/SOFTWARE/TOOLS/PROGRAMMING/petsc/src/ksp/ksp/examples/tutorials/ex10 -f0 A_and_rhs.bin  -trans -pc_type asm
Number of iterations = 10000
Residual norm 1.30198

$ mpirun -np 4 /gpfs/fs1/aero/SOFTWARE/TOOLS/PROGRAMMING/petsc/src/ksp/ksp/examples/tutorials/ex10 -f0 A_and_rhs.bin  -trans -pc_type asm
Number of iterations = 207
Residual norm 0.000555849


---------------

$ mpirun -np 4 /gpfs/fs1/aero/SOFTWARE/TOOLS/PROGRAMMING/petsc/src/ksp/ksp/examples/tutorials/ex10 -f0 A_and_rhs.bin  -trans -pc_type gasm
Number of iterations = 297
Residual norm 0.000600143

$ mpirun -np 4 /gpfs/fs1/aero/SOFTWARE/TOOLS/PROGRAMMING/petsc/src/ksp/ksp/examples/tutorials/ex10 -f0 A_and_rhs.bin  -trans -pc_type gasm
Number of iterations = 297
Residual norm 0.000600143



Antoine DeBlois
Specialiste ingenierie, MDO lead / Engineering Specialist, MDO lead
Aéronautique / Aerospace
514-855-5001, x 50862
antoine.deblois at aero.bombardier.com<mailto:antoine.deblois at aero.bombardier.com>

2351 Blvd Alfred-Nobel
Montreal, Qc
H4S 1A9

[Description : Description : http://signatures.ca.aero.bombardier.net/eom_logo_164x39_fr.jpg]
CONFIDENTIALITY NOTICE - This communication may contain privileged or confidential information.
If you are not the intended recipient or received this communication by error, please notify the sender
and delete the message without copying

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20140922/a3f552b2/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.jpg
Type: image/jpeg
Size: 4648 bytes
Desc: image001.jpg
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20140922/a3f552b2/attachment-0001.jpg>


More information about the petsc-users mailing list