Re: [petsc-users] Call to PETSc functions way higher when using lower number of processors

[Jose A. Abell M.]

On Wed, Jul 1, 2015 at 4:34 PM, Barry Smith <bsm...@mcs.anl.gov> wrote:

>   I suggest you run the following experiment; run with ONE process but use
> -pc_type bjacobi -sub_pc_type ilu -pc_bjacobi_blocks <blocks> where you use
> for <blocks> 1 up to 24 and then get the number of iterations needed for
> each (don't worry about the time it takes, this is done for understanding
> of the convergence). Send the table of
>

​Hi Barry,

Thanks for the feedback! I switched to SuperLU_dist (-ksp_type preonly
-pc_type lu -pc_factor_mat_solver_package superlu_dist) and saw immediate
speedup for all number of processors. I'll try your experiment and see what
comes up.

Thanks!​


--

José Abell
PhD Candidate
Computational Geomechanics Group
Dept. of Civil and Environmental Engineering
UC Davis