[petsc-users] Variable wall times

Phanisri Pradeep Pratapa ppratapa at gatech.edu
Mon Feb 22 12:44:33 CST 2016


I am trying to compare run times for a function in petsc using ksp solve.
When I run the code in parallel on a single node (64 processors), it seems
to take different wall time for each run, although run on exactly same
processors. For example one time it could take 10 sec where as the same run
next time could take 20 sec. (This happens for other choices of processors
too, not just 64)

I am solving a linear system using gmres. I have attached four output files
for the same inputs and run on same processors. You can see the
differences. It seems like the variability comes from each processor taking
different time to complete the MatMult and KSPGMRESOrthog (you can see
profiling output from petsc in the attached files). Please let me know if
you have any idea on how I can overcome this problem.

Thank you,


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20160222/80a47136/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: output_files.zip
Type: application/zip
Size: 13493 bytes
Desc: not available
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20160222/80a47136/attachment.zip>

More information about the petsc-users mailing list