On Sat, Mar 6, 2010 at 3:41 PM, <span dir="ltr"><<a href="mailto:hxie@umn.edu">hxie@umn.edu</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
Hi,<br>
<br>
I am solving incompressible navier-stokes equations. The code implements the Newton iteration and just uses the KSP solver for each iteration. It takes 14 nonlinear iterations (~8min) to converge using the default solver in PETSc. But it takes 6 nonlinear iterations (~30min) using boomeramg from hypre. The amg taks less iterations but takes too much cputime. Is this normal? One friend told me the reason may be that I did not call the destroy function to free the memory in each Newton iteration (I called PetscFinalize after all the Newton iterations). How can I call the PCDestroy_HYPRE from a fortran code? Will there be significant difference in cputime to call amg directly from hypre? Thanks for any help.<br>
</blockquote><div><br></div><div>This has nothing to do with the memory.BoomerAMG does a lot of processing.</div><div><br></div><div> Matt</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
Bests,<br>
Hui</blockquote></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<br>