[O-MPI users] Re: [Beowulf] Alternative to MPI ABI
landman at scalableinformatics.com
Sun Apr 3 18:06:42 PDT 2005
Mark Hahn wrote:
>>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.
> wonderful! yes: ABI standards are good and proprietary
> implementations (which inherently provide only negative
> definitions of support) are bad.
Er... standards are good. Proprietary implementations ... hmmm... don't
know where you were pulling that one from.
I am complaining about being forced to have (no offence meant Jeff,
William, et al, good work BTW) LAM 7.0.3, 7.0.6, 7.1.1; MPICH 22.214.171.124,
126.96.36.199, 1.2.6.x) on one machine, with the associated problems with
maintenance and configuration that this brings (environment variables,
Last I checked, MPICH and LAM are not proprietary. They do however have
different ABIs, though they expose nearly the same API. Moreover, a
code compiled with LAM 7.0.3 does not seem to like being run using the
LAM 7.1.1 tools (and if there is a way to do this, please, someone, let
me know offline!!!). If this is supposed to work, I'd be happy to file
a bug report.
> after all, the real appeal is that N MPI implementation only need to test
> their own conformity to the standard, and M applications test their
> conformity. ie, N+M tests, rather than N*M without an ABI. this
> assumes that the ABI/standard is broad enough, of course!
If they interoperate ... (sigh)
> first, it's worth asking whether there is something to be lost
> by going to an ABI? yes, dynamic linking imposes some overhead -
> I have to wonder whether some of the higher-performing interconnects
> (SGI/Cray/Quadrics/Pathscale) are low-latency enough to worry about
> indirect library calls blowing the pipeline.
The indirection should not cost a great deal (well under single byte one
way transfer latency time) or the platform vendor has other issues they
have to deal with. It could though.
Imagine, to do messaging on future fabrics, we might need an RTOS :(
Joseph Landman, Ph.D
Founder and CEO
Scalable Informatics LLC,
email: landman at scalableinformatics.com
web : http://www.scalableinformatics.com
phone: +1 734 786 8423
fax : +1 734 786 8452
cell : +1 734 612 4615
More information about the Beowulf