[petsc-users] Scalability of PETSc

[Barry Smith]

   Bibrak,

   Start with http://www.mcs.anl.gov/petsc/documentation/faq.html#computers

    Then http://www.mcs.anl.gov/petsc/documentation/faq.html#log-summary

     and http://www.mcs.anl.gov/petsc/documentation/faq.html#slowerparallel

    Barry

On Apr 2, 2012, at 7:37 AM, Bibrak Qamar wrote:

> Hello all,
> 
> I am using PETSc to solve linear equation using CG. I am not getting good scalability (using the MPI implementation, not GPUs). One reason can be that may be for my platform I haven't tuned the library. Other reason may be the nature of CG, that it is communication bound, at each iteration the vector (search direction) has to be communicated. 
> 
> Where can I find statistics of the scalability of CG PETSc (if there is any, people may have reported)?
> 
> Or can anyone give a number on how well CG PETSc scales on their environment. Lets take the simple example provided in the PETSc KSP tutorials - KSP Laplacian, 2d (ex2.c) for mesh dim of 
> n = 150000
> m = 100
> 
> 
> Thanks
> Bibrak