[petsc-dev] Kokkos/Crusher perforance

[Mark Adams]

>
>
>
> > VecPointwiseMult     201 1.0 1.0471e-02 1.1 3.09e+08 1.0 0.0e+00 0.0e+00
> 0.0e+00  0  1  0  0  0   1  1  0  0  0 235882   290088      0 0.00e+00    0
> 0.00e+00 100
> > VecScatterBegin      200 1.0 1.8458e-01 1.1 0.00e+00 0.0 1.1e+04 6.6e+04
> 1.0e+00  2  0 99 79  0  19  0100100  0     0       0      1 2.04e-04    0
> 0.00e+00  0
> > VecScatterEnd        200 1.0 1.9007e-02 3.7 0.00e+00 0.0 0.0e+00 0.0e+00
> 0.0e+00  0  0  0  0  0   1  0  0  0  0     0       0      0 0.00e+00    0
> 0.00e+00  0
>
> I'm curious how these change with problem size. (To what extent are we
> latency vs bandwidth limited?)
>
>
I am getting a segv in ex13 now, a Kokkos view in Plex, but will do scaling
tests when I get it going again.
(trying to get GAMG scaling for Todd by the 3rd)



> > SFSetUp                1 1.0 1.3015e-03 1.3 0.00e+00 0.0 1.1e+02 1.7e+04
> 1.0e+00  0  0  1  0  0   0  0  1  0  0     0       0      0 0.00e+00    0
> 0.00e+00  0
> > SFPack               200 1.0 1.7309e-01 1.1 0.00e+00 0.0 0.0e+00 0.0e+00
> 0.0e+00  2  0  0  0  0  18  0  0  0  0     0       0      1 2.04e-04    0
> 0.00e+00  0
> > SFUnpack             200 1.0 2.3165e-05 1.4 0.00e+00 0.0 0.0e+00 0.0e+00
> 0.0e+00  0  0  0  0  0   0  0  0  0  0     0       0      0 0.00e+00    0
> 0.00e+00  0
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20220125/e47e862f/attachment.html>