Author Topic: Linux Magazine: Why MPI is inadequate  (Read 3529 times)

Offline contingencyplan

  • Villain
  • Moderator
  • Admiral
  • *****
  • Posts: 977
  • Karma: +1/-0
  • Must I sin once, and repent forever?
    • View Profile
    • My Blog
Linux Magazine: Why MPI is inadequate
« on: September 18, 2006, 12:52:53 pm »
One of the things I mentioned in the RaptorNL initial post was that I am considering (eventually!) putting support for MPI (or something similar, maybe even home-grown) into the library. I ran across this article on Linux Magazine's website (registration required) that basically goes into why MPI is inadequate. I'd like more information on the socket-based solutions and clustering software referenced in the article, but I'm sure there's articles about that stuff elsewhere on the site (and definitely on other websites out there). At any rate, here ya go - let's start a discussion about this. I figure Precursors especially will need support for multiprocessing (not just multithreading), so this is something we need to give more than due consideration to.
We've all heard that a million monkeys banging on a million typewriters will eventually reproduce the entire works of Shakespeare. Now, thanks to the Internet, we know this is not true. 
    ~Robert Wilensky

It is not bigotry to be certain we are right; but it is bigotry to be unable to imagine how we might possibly have gone wrong.
    ~GK Chesterton

Men never do evil so completely and cheerfully as when they do it from a religious conviction.
    ~Blaise Pascal