<div dir="ltr">On Wed, Apr 10, 2013 at 10:04 AM, Chris Kees <span dir="ltr"><<a href="mailto:cekees@gmail.com" target="_blank">cekees@gmail.com</a>></span> wrote:<br><div class="gmail_extra"><div class="gmail_quote">
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">Hi guys,<br>
<br>
Could somebody point me to some examples you guys routinely use for<br>
weak and strong scaling studies (maybe even with scripts, option<br>
files, or prior results on recent hardware)? I'm thinking of 3D<br>
Poisson with finite differences and geometric multigrid or something<br>
like that.<br></blockquote><div><br></div><div style>I am trying to write this stuff down, but there is not much written right now.</div><div style>Let's start at the beginning. SNES ex5 is the simplest example that can be</div>
<div style>used for this I think. It is 2D Poisson (actually Bratu). Its really easy to get</div><div style>weak scaling by adjusting the grid size using</div><div style><br></div><div style> -da_grid_x <m> -da_grid_y <n></div>
<div style><br></div><div style>You can turn on MG using</div><div style><br></div><div style> -pc_type mg -pc_mg_levels <n></div><div style><br></div><div style>although the slightly non-intuitive thing is that then the grid size you input is</div>
<div style>for the coarse grid.</div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">
We've been trying to work toward scaling studies of the field split<br>
and Schur complement preconditioners for our multiphase flow solvers,<br>
but I'm realizing that we need to do more thorough testing of the<br>
petsc installation itself and make sure we're using timing/profiling<br>
"best practices" and such.<br>
<br>
We are using petsc-dev on the hardware below. I promise to quit using<br>
petsc-dev as soon as the next release comes out:) Several versions of<br>
PETSc are also installed by the system maintainers, but my sense is<br>
that there is very little testing done on any of the installations.<br></blockquote><div><br></div><div style>I think using petsc-dev is the "right thing", and it is now much much more</div><div style>stable (the 'master' branch on BitBucket).</div>
<div style><br></div><div style> Thanks,</div><div style><br></div><div style> Matt</div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">
<a href="http://www.erdc.hpc.mil/hardware/index.html" target="_blank">http://www.erdc.hpc.mil/hardware/index.html</a><br>
<span class=""><font color="#888888"><br>
Chris<br>
</font></span></blockquote></div><br><br clear="all"><div><br></div>-- <br>What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>
-- Norbert Wiener
</div></div>