[Chicago] MPI in Python?

Chris McAvoy chris.mcavoy at gmail.com
Wed Feb 27 16:21:38 CET 2008

On Wed, Feb 27, 2008 at 8:31 AM,  <skip at pobox.com> wrote:
> Is anyone using MPI within Python to harness the compute power of multiple
>  CPU machines?

No, but I love your use of the phrase "harness the compute power."  Well played.

I have been playing around with IPython1, the poorly named
very-easy-to-use clustering thing:

Peter Skomoroch (http://www.datawrangling.com/) is presenting at Pycon
on clustering with MPI on Amazon EC2.  I had proposed a similar talk,
just replacing MPI with IPython1.  We were going to collaborate, but I
ended up dropping out.  Peter is still going to (I think) touch on
some IPython1 stuff in his talk.

As an aside, I'm (very slowly) reading the O'Reilly book "Collective
Intelligence" which explains pretty clearly a handful of common data
mining techniques.  I'd like to start playing around with them with
IPython1 applying them to the Netflix data set, probably after Pycon.
I planned on working in my public subversion repository, to be...you
know..public.  I'll post links if I ever get anything done.  Again,
probably after Pycon.


More information about the Chicago mailing list