Cisco Blogs

Cisco Blog > High Performance Computing Networking

Exascale: it’s not just the (networking) hardware

January 24, 2011 at 7:45 am PST

Many in the HPC research community are starting to work on “exascale” these days — the ability to do 10^18 floating point operations per second.  Exascale is such a difficult problem that it will require new technologies in many different areas before it can become a reality.  Case in point is this entry at Inside HPC today entitled, “InfiniBand Charts Course to Exascale“.

It cites The Exascale Report and a blog entry by Lloyd Dickman at the IBTA about their course going forward.  It’s a good read — Lloyd’s a smart, thoughtful guy.

That being said, there’s a key piece missing from the discussion: the (networking) software.  More specifically: the current OpenFabrics Verbs API abstractions are (probably) unsuitable for exascale, a fact that Fab Tillier (Microsoft) and I presented at the OpenFabrics workshop in Sonoma last year (1up, 2up).

Read More »

Tags: , , , , ,

Community-contributed Perl and Python bindings for hwloc

January 22, 2011 at 7:30 am PST

I love open source communities.

Two hwloc community members have taken it upon themselves to provide high-quality native language bindings for Perl and Python.  There’s active work going on, and discussions occurring between the hwloc core developers and these language providers in order to provide good abstractions, functionality, and experience.

  • The Perl CPAN module is being developed by Bernd Kallies: you can download it here (I linked to the directory rather than a specific tarball because he keeps putting up new versions).
  • The Python bindings are being developed by Guy Streeter (at Red Hat); his git repository is available here.

Read More »

Tags: ,

Why MPI?

January 7, 2011 at 3:15 pm PST

It’s the beginning of a new year, so let’s take a step back and talk about what MPI is and why it is a Good Thing.

I’m periodically asked what exactly MPI is.  Those asking cover many different biases: network administrators, systems programmers, application programmers, web developers, server and network hardware designers, … the list goes on.  Most have typically heard about this “MPI” thing as part of “high performance computing” (HPC), and think that it’s some kind of parallel programming model.

Technically, it’s not.  MPI — or, more specifically, message passing — implies a class of parallel programming models.  But at its heart, MPI is about simplified inter-process communication (IPC).

Read More »

Tags: ,

The Graph 500

December 17, 2010 at 5:00 am PST

Did you hear about the Graph 500 at SC’10?  You might not have.  It got some fanfare, but other press releases probably drowned it out.

Even though it’s a brand new “yet another list”, it’s worth discussing because it’s officially a Good Idea.  Here’s what Rich Murphy, Official Chief Graph 500 Cat Herder (ok, I might have made up that title), tells me about it:

Basically, what we’re trying to do is create a complementary measure to Linpack for data intensive problems.  A lot of us on the steering committee believe that these kinds of problems will dominate high performance computing over the next decade.  We’ve given some “business areas” as examples of these kinds of applications: cybersecurity, medical informatics, data enrichment, social networks, and symbolic networks.  These basically exist to support the assertion that this could be huge someday.

+1 on what he says.

Read More »

Tags: ,

Hardware Locality (hwloc) v1.1 released

December 16, 2010 at 4:20 pm PST

I’m very pleased to announce that we just released Hardware Locality (hwloc) version 1.1.  Woo hoo!

There’s bunches of new stuff in hwloc 1.1:

  • A memory binding interface is the Big New Feature.  It’s available in both the C API and via command line options to tools such as hwloc-bind.
  • We improved lotopo’s logical vs. physical ID numbering.  Logical numbers are now all prefixed with “L#”; physical numbers are prefixed with “P#”.  That’s that, then.
  • “cpusets” are now “bitmaps”, and now have no maximum size; they’re dynamically allocated (especially for machines with huge core counts).
  • Arbitrary key=value caching is available on all objects.

…more after the break.

Read More »

Tags: , ,