[Beowulf] The Case for an MPI ABI

Rusty Lusk lusk at mcs.anl.gov
Tue Feb 22 09:06:05 PST 2005


From: Greg Lindahl <lindahl at pathscale.com>
Subject: [Beowulf] The Case for an MPI ABI
Date: Mon, 21 Feb 2005 21:08:24 -0800
> This talk:
> 
> http://www.openib.org/docs/oib_wkshp_022005/mpi-abi-pathscale-lindahl.pdf
> 
> mostly talks about why we need an ABI, who wins and loses as a result
> of having one, and the pieces that could be in it. Please give it a
> look.

One piece you include is the replacement of the non-portable mpirun with
an mpistart with standard arguments.  You might note that the MPI-2
forum addressed this issue with the specification of an mpiexec with
standard arguments.  MPICH2 implements it.

Rusty





More information about the Beowulf mailing list