On Wed, Jun 03, 2009 at 07:24:23AM -0500, Jeff Squyres wrote: > I'll state right up front that I'm in favor of removing MPI from > OFED. :-) > > > On Jun 2, 2009, at 3:30 PM, Tziporet Koren wrote: > > > Here are some of the reasons for keeping MPI in or out of OFED. > > > > Main reasons to take MPI out of OFED: > > - MPI is not developed under OFA > > - Need to synchronize between different projects > > > --> This point alone is hugely difficult. OFED has been held up > multiple times because of MPI release delays. I would like to stress > that the open source MPI implementations are entirely different > software projects with minimal overlap of personnel between MPI and > OFED. Also, OFED is but one of the stacks that the MPI's support.
On our systems we currently have 21 (just counted them) different versions of of MPI installed. And we are thinking about adding OpenMPI into the mix. In addition, there are 24 different compilers, 29 scientific libraries, 8 tool sets, 5 debuggers and 4 vis tools. This, all in addition to what comes in the SLES release. Point is - adding another MPI module is orthogonal to updating OFED. I don't see a good reason to have an artificial release dependency. Any serious regression test for OFED needs to have been tested against much more than OpenMPI. Its far more important to improve the OFED release and not break existing code/ULP than to get another MPI released at the same time. The only possible counter I can see is if OpenMPI is better able to drive feature enhancement/development in OFED than say mvapich. DK? In my opinion, anything that slows the OFED release is a bad thing. Of course, that does not mean we are willing to do more ;) debugging in the field. bob ------------------------------------------------------------------------- Robert B. Ciotti Supercomputing Systems Lead NASA Advanced Supercomputing (NAS) Division TEL (650) 604-4408 NASA Ames Research Center FAX (650) 604-4377 Moffett Field, CA 94035-1000 [email protected] ------------------------------------------------------------------------- _______________________________________________ general mailing list [email protected] http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
