PySPH with MPI - example

[Daniel Cordier]

Hi Everyone,

   I am working with bastien and I have a similar problem : MPI simulations (dambreak-like)

with a large number of particles and a number of CPU > 60 crash, could someone share

and example (pysph source & slurm script) of such simulation running  without problem

on a super-computer? Many thanks in advance for any help

Daniel

[Prabhu Ramachandran]

Hi everyone,

Sorry I have been swamped to look into this. Are you doing pure MPI or are you doing a combination of OpenMP + MPI?  We did run a case with 80 MPI procs for the paper but unfortunately we have not been using the MPI code enough recently due to very long wait times on our own cluster currently.

Do you have a simple test case? We could try to run it and see.

Regards,
Prabhu

--
You received this message because you are subscribed to the Google Groups "pysph-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pysph-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pysph-users/b1cae89c-6079-4d1a-bedb-55463f829359n%40googlegroups.com.