The preconditioner also matters. If you are using a one-level preconditioner, the 3D partition can cut the iteration count by a factor of approximately the cube root of the number of processes relative to the 1D partition.<div>
<br></div><div>There are many factors that can limit the performance of your code. For performance questions, always send the output of -log_summary.<br><br><div class="gmail_quote">On Tue, Sep 4, 2012 at 11:55 PM, TAY wee-beng <span dir="ltr"><<a href="mailto:zonexo@gmail.com" target="_blank">zonexo@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="im">On 4/9/2012 1:54 PM, Aron Ahmadia wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
There are a lot of other factors at play here, including how much time<br>
you are spending working on the code, how far you're trying to scale<br>
the algorithm and how much communication dominates your problem for<br>
the architectures and algorithms you are using. Unless you are trying<br>
to take this code to thousands of processors I probably wouldn't worry<br>
about it.<br>
<br>
A<br>
</blockquote>
<br></div>
Ok thanks for the suggestions Aron. I'll look at others parts for optimization.<div class="HOEnZb"><div class="h5"><br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
On Tue, Sep 4, 2012 at 11:17 AM, TAY wee-beng <<a href="mailto:zonexo@gmail.com" target="_blank">zonexo@gmail.com</a>> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
On 4/9/2012 11:11 AM, Aron Ahmadia wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
This doesn't strike me as a particularly large problem. I'm not sure<br>
it's worth doing unless you are going to be looking at more unknowns<br>
in the future.<br>
<br>
A<br>
</blockquote>
Hi Aron,<br>
<br>
It will increase to 500x500x1200 or more. In that case, do you think it's<br>
worth it?<br>
<br>
Thanks!<br>
<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
On Tue, Sep 4, 2012 at 10:08 AM, TAY wee-beng <<a href="mailto:zonexo@gmail.com" target="_blank">zonexo@gmail.com</a>> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Hi,<br>
<br>
My Fortran CFD code is currently partitioned in the z direction. Total<br>
grid<br>
size is around is 153x248x620. Hence depending on the no. of procs, the z<br>
direction 620 is partitioned. The grid size changes but the ratio is<br>
around<br>
there.<br>
<br>
Parititoning in 3 directions was initially too complex for me. However,<br>
it<br>
seems to be much simplified with the use of DM. However, there's still a<br>
lot<br>
of work to be done to make it working.<br>
<br>
I'm wondering how much improvement in speed/efficiency will I get, if I<br>
partition from 1 direction to 3 directions. Is it worth the effort?<br>
<br>
--<br>
Yours sincerely,<br>
<br>
TAY wee-beng<br>
<br>
</blockquote></blockquote></blockquote></blockquote>
<br>
</div></div></blockquote></div><br></div>