[Chicago] Parallelism, or Chris is a big thinker

[Cosmin Stejerean]

Well, along those lines IPython1 should bring some really nice capabilites
in this field by being able to leverage MPI to distribute work from one
IPython session to several slaves and collect the results so you can
leverage your dual core machine or multiple machines to accomplish tasks
faster (http://ipython.scipy.org/moin/Parallel_Computing)

- Cosmin

On 6/6/07, Chris McAvoy <chris.mcavoy at gmail.com> wrote:
>
> Remember yesterday when we were all talking about concurrency and
> stuff?  I guess a bunch of big thinkers were too:
>
> http://radar.oreilly.com/archives/2007/06/googles_folding.html
>
> I ended up solving my teeny tiny concurrency problem by opening up two
> ipython sessions and running my jobs in batches.  It was cheap, but it
> sped up my job by *2, and I got to see my dual core machine go to 100%
> rather than 50%.
>
> Next week I'll show anyone that's interested.  It really bugged me the
> first time I opened up my mac processor monitor thingee and saw Python
> 100% CPU, but my actual CPU at 50%.  Rather than get into fork, I took
> the cheap way by running in batches...but hey...cheap works.
>
> Chris
> _______________________________________________
> Chicago mailing list
> Chicago at python.org
> http://mail.python.org/mailman/listinfo/chicago
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/chicago/attachments/20070606/76727471/attachment.html