The problem is small enough that you might be able to use MUMPS.<br><br> Matt<br><br><div class="gmail_quote">On Fri, Jun 12, 2009 at 9:31 AM, Lisandro Dalcin <span dir="ltr"><<a href="mailto:dalcinl@gmail.com">dalcinl@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><div class="im">On Fri, Jun 12, 2009 at 11:13 AM, Christian<br>
Klettner<<a href="mailto:christian.klettner@ucl.ac.uk">christian.klettner@ucl.ac.uk</a>> wrote:<br>
> Sorry that I sent this twice. No subject in the first one.<br>
><br>
> Dear PETSc Team,<br>
> I am writing a CFD finite element code in C. From the discretization of<br>
> the governing equations I have to solve a Poisson type equation which is<br>
> really killing my performance. Which solver/preconditioner from PETSc or<br>
> any external packages would you recommend? The size of my problem is from<br>
> ~30000-100000 DOF per core. What kind of performance would I be able to<br>
> expect with this solver/preconditioner?<br>
<br>
</div>I would suggest KSPCG. As preconditioner I would use ML or<br>
HYPRE/BoomerAMG (both are external packages)<br>
<div class="im"><br>
> I am using a 2*quad core 2.3 GHz Opteron. I have decomposed the domain<br>
> with Parmetis. The mesh is unstructured.<br>
> Also, I am writing a code which studies free surface phenomena so the mesh<br>
> is continually changing. Does this matter when choosing a<br>
> solver/preconditioner? My left hand side matrix (A in Ax=b) does not<br>
> change in time.<br>
<br>
</div>ML has a faster setup that BoomerAMG, but the convergence is a bit<br>
slower. If your A matrix do not change, then likely BoomerAMG will be<br>
better for you. In any case, you can try both: just build PETSc with<br>
both packages, then you can change the preconditioner by just passing<br>
a command line option.<br>
<div class="im"><br>
><br>
> Best regards and thank you in advance,<br>
> Christian Klettner<br>
><br>
<br>
</div>Disclaimer: the convergence of multigrid preconditioners depends a lot<br>
on your actual problem. What I've suggested is just my limited<br>
experience in a few problems I've run solving electric potentials.<br>
<font color="#888888"><br>
<br>
--<br>
Lisandro Dalcín<br>
---------------<br>
Centro Internacional de Métodos Computacionales en Ingeniería (CIMEC)<br>
Instituto de Desarrollo Tecnológico para la Industria Química (INTEC)<br>
Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET)<br>
PTLC - Güemes 3450, (3000) Santa Fe, Argentina<br>
Tel/Fax: +54-(0)342-451.1594<br>
</font></blockquote></div><br><br clear="all"><br>-- <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<br>