[Beowulf] Re: Povray + MPI (award at andorra.ad)

David Mathog mathog at mendel.bio.caltech.edu
Tue Sep 20 13:40:45 PDT 2005

> From: award at andorra.ad

> Could anybody give me a pointer to either a compatible MPI Patch + Povray
> source, or a compiled version for RedHat 9 + Gcc 3.2.2 ? Any >= 3.0
> version would be fine.

This probably isn't what you want to hear, but if you N nodes
and M rendering jobs it will run faster if you
just distribute those M normal Povray jobs across those N nodes
(with the queue system of your choice) than if you build an
MPI or PVM version of POVRAY and then run the M jobs
sequentially through that program.

At least that's the way it was when I last looked at it several years

This is true so long as you DON'T have the remote jobs writing povray
output continuously back to your NFS server.  It works better to
direct to /tmp and then copy in one fell swoop back to the final
central directory.


David Mathog
mathog at caltech.edu
Manager, Sequence Analysis Facility, Biology Division, Caltech

More information about the Beowulf mailing list