[openib-general] mpi on infiniband and openib

Mon Mar 29 14:45:29 PST 2004

Troy, 

You must have received a 2-line patch (related to memory
registration/deregistration in MVAPICH2) from one of my students a
short while ago. You can try that and let me know whether the problem
is solved or not.

Please note that the MVAPICH2 release (0.5a) is an `alpha' release
based on MPICH2 0.94b. I do not think we have touched anything related
to MPI communicators.

If the problem persists, you can let us know the exact details of the
errors you are seeing and we will be happy to take a look at it.

I will also suggest you to try GAMESS with MVAPICH. Based on our
earlier conversation, I think it was running successfully with
MVAPICH. Please let me know the status.

Thanks, 

DK

> On Mon, Mar 29, 2004 at 02:31:27PM -0500, Dhabaleswar Panda wrote:
> > > 
> > > What mpi implementations are available in the IB world?  Specifically, I'd 
> > > like to compile up IOR 
> > > (http://www.llnl.gov/asci/purple/benchmarks/limited/ior/) and get it running 
> > > to test out a filesystem.  Any help with a starting point would be greatly 
> > > appreciated.  Thanks...
> > 
> > MVAPICH and MVAPICH2 (MPI-2 version) from OSU are available for the IB
> > world and are being used at many places world-wide. The current
> > `infiniband-support-2004-02-26.tar.bz2' file in the openib downloads
> > includes a copy of MVAPICH.
> 
> Has anyone tested this with applications using MPI communicators? We
> (myself and Brett Bode at Ames Laboratory) compiled MVIAPICH-2 recently
> and tried to run GAMESS, and it works on 2 processors, but any more than
> that and it gets the wrong answer.
> 

-- 
To unsubscribe send an email with subject unsubscribe to openib-general at openib.org.
Please contact moderator at openib.org for questions.