[Beowulf] The Case for an MPI ABI

[Joachim Worringen]

Ashley Pittman wrote:
> It's not about competence, it's about time and effort spent.  I wouldn't
> need to compile every application myself if there was a ABI.  It's not a
> particularly difficult thing to do, it just going through the hoops of
> doing it every time you need an application.  The ability to install a
> cluster and type '[apt-get|yum] install pmb' would be a truly wonderful
> thing indeed.

My experience is that such sites want to use their (expensive, 
commercial, Fortran) compiler to optimize the binaries to their 
platform. In some cases, this requires source-code changes (parameters) 
anyway. It's not about PMB.

> You also make the assumption that it's the high-performance vendors who
> do things differently, I don't believe this is the case.  Quadrics for
> example (my employer) happen to use whatever ABI MPICH (1.2.x) provides
[...]

I can't see where I made this assumption. Indeed, most interconnect 
vendors (Quadrics, various Infiniband, Myrinet, ...) happily plug their 
low-level stuff into the latest MPICH and are done. So, it's most often 
the cross-interconnect MPI vendors which create their own ABI for some 
reason. Other cases are vendors (like us) who started to provide MPI-2 
when there was no open-source MPI-2 around. We had to do our own 
definitions then.

But this doesn't really matter after all; what matters is if there are 
enough parties to take part in this effort, and to understand the 
related issues as much as possible and as early as possible.

  Joachim

-- 
Joachim Worringen - NEC C&C research lab St.Augustin
fon +49-2241-9252.20 - fax .99 - http://www.ccrl-nece.de