Cisco Blogs


Cisco Blog > High Performance Computing Networking

What’s Next for MPI?

February 18, 2014 at 3:40 am PST

MPI-3 has been out for over a year and a half.  MPICH supports all of the mandatory MPI-3 behavior and some of its optional semantics.  Open MPI supports all of MPI-3 except the new one-sided semantics.  New functionality is becoming mature in both, and that maturity is trickling down to the implementations that are derived from those two.

So what’s next?  Are we done?

Far from it.

Read More »

Tags: ,

Open MPI 1.7.4 released!

February 7, 2014 at 6:36 am PST

Open MPI logoIt took us longer than we intended, but we finally released Open MPI v1.7.4.  

Woo hoo!  (we got nice coverage from El Reg, too)

This is a monster release; it represents hundreds (thousands? millions?) of person-hours of work.  Consider this a ginormous “thank you!” to the entire Open MPI community! Special thanks goes to Paul Hargrove of Lawrence Berkeley Labs for his hero efforts in regression testing throughout December and January.

Read More »

Tags: , ,

More Network Locality (Netloc) progress

January 31, 2014 at 9:11 am PST

We announced the Network locality project at SC’13, and generated a LOT of interest (far more than I even anticipated!).  As a refresher, here’s a link to a a blog entry we wrote about Netloc back in November.

There is still much work to be done; we’re actively continuing work in multiple areas:
Read More »

Tags: , , ,

InsideHPC podcast: MPI collaboration with OpenFabrics

January 30, 2014 at 11:03 am PST

In my last blog post, I described a new collaboration between the MPI community and the OpenFabrics verbs community.

The collaboration started with the OpenFrameworks group asking the MPI community to list its requirements for a lower layer network API to the OpenFabrics OpenFrameworks working group.

In that last blog post, I posted an abbreviated set of slides with the first part of this MPI community feedback.  Since then, I have finished collating the feedback from the MPI community and have assembled a full set of slides.

Read More »

Tags: , ,

A fun thing happened on the way to the OpenFrameworks discussion today…

January 21, 2014 at 12:11 pm PST

A few months ago, Sean Hefty from Intel started an effort to design a new low-level network API to replace libibverbs.

That is, it’s not libibverbs 2.0 — it’s a new API that aims to both expand the scope of what libibverbs did, and also to address many of its much-criticized shortcomings.  Sean and Paul Grun even held a BoF about the proposed “libfabric” at SC’13.

Design discussions have been progressing within the “OpenFrameworks” working group in OpenFabrics. Progress has mainly occurred in the form of emails and weekly conference calls. Two weeks ago, Sean asked me to gather feedback about what an MPI implementation would want from a new underlying network API, and present that feedback to the libfabric group.

So I asked the MPI community.

Read More »

Tags: , ,