[Chicago] MPI in Python?

skip at pobox.com skip at pobox.com
Wed Feb 27 19:06:10 CET 2008

    Massimo> If you need parallelization ,you care about speed. If you care
    Massimo> about speed use C/C++.

We do already, only indirectly via packages like numpy.

    Massimo> You use threads on shared memory machines (like a SG), you use
    Massimo> MPI on distributed memory machines (like a cluster). There is
    Massimo> really no choice.

I effectively have both.  Threads gain nothing from within Python because of
the interpreter lock.  I don't need to worry too much about buying more
hardware, but I do want to use it somewhat efficiently.  If I have a
computer with 16 cores I need to use multiple processes to make reasonably
effective use of it from Python.


More information about the Chicago mailing list