I also did this on my software and it shows terrific performance in serial.<br>My problem may come from the default distribution of the matrix in PETSc : N rows on proc 0, N rows on proc 1, ....<br><br>How did you distribute the matrix in your program?
<br><br>Nicolas<br><br><div><span class="gmail_quote">2008/1/16, Sanjay Govindjee <<a href="mailto:sanjay@ce.berkeley.edu">sanjay@ce.berkeley.edu</a>>:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
I have not tried this with any of the examples in the distribution only<br>with my own program.<br><br>The steps are relatively straightforward but you must know them already<br>since you have it<br>working in serial. Nonetheless here is the order of call I have:
<br><br>call MatCreateMPIBAIJ(... ! must be blocked<br>call MatAssemblyBegin (... ! Assemble the matrix<br>call MatAssemblyEnd (...<br><br>call VecAssemblyBegin (... !Assemble the RHS
<br>call VecAssemblyEnd (...<br><br>call KSPCreate (...<br>call KSPSetOperators (...<br>call KSPGetPC (...<br>call PCSetCoordinates(... ! I'm not fully certain<br>but your problem may have to be defined over R^3 for this to work well
<br>call KSPSetFromOptions(...<br><br>call KSPSolve (...<br><br><br>Nicolas Tardieu wrote:<br>> Hi Sanjay,<br>><br>> Have you used it on simple examples of the PETSc distribution?<br>><br>> Nicolas<br>
><br>> 2008/1/16, Sanjay Govindjee <<a href="mailto:sanjay@ce.berkeley.edu">sanjay@ce.berkeley.edu</a><br>> <mailto:<a href="mailto:sanjay@ce.berkeley.edu">sanjay@ce.berkeley.edu</a>>>:<br>><br>> We have used it extensively with good success in the past.
<br>> -sanjay<br>><br>> Nicolas Tardieu wrote:<br>> > Hi everyone,<br>> ><br>> > I have been using Prometheus for scalar and vector problems with<br>> great<br>> > success on sequential computers.
<br>> > But when trying to use it in parallel, Prometheus crashes.<br>> > Then I tried it on some PETSc tests<br>> (src/ksp/ksp/examples/tests/ex3.c<br>> > for instance) on parallel and they also crash...
<br>> > Does anyone have experienced Prometheus in parallel?<br>> ><br>> > Thanks,<br>> ><br>> > Nicolas<br>> ><br>><br>><br><br></blockquote></div><br>