Jeff Squyres wrote:
I have a followup question:
Who, exactly, wants an MPI ABI? I have seen a vocal few voice their
opinions (both for and against). But these are not representative of
<hand shoots skyward>
Me... please, please.... really... I don't want 7 MPI implementations on
my customers clusters anymore. I really want to dynamically compile
against an MPI spec (really have the commercial vendors do this), and
then at run time link them to the correct transport driver library.
Sorta like MVIAPICH.
If there is an ABI then we have a fighting chance at focusing on the
applications and not the ever-so-slightly-strange version of whichever
flavor of MPI that they chose to use.
--
Joseph Landman, Ph.D
Founder and CEO
Scalable Informatics LLC,
email: land...@scalableinformatics.com
web : http://www.scalableinformatics.com
phone: +1 734 786 8423
fax : +1 734 786 8452
cell : +1 734 612 4615