MPI is really independent of places/futures because it is process based. We could do an MPI-like or even MPI compatible Racket library.
For CUDA, it is easier, because Racket already has a featureful OpenCL implementation: http://planet.plt-scheme.org/display.ss?package=opencl.plt&owner=jaymccarthy In fact, the development version is getting much better: faster & more compatible: http://github.com/jeapostrophe/opencl Jay On Sun, Oct 3, 2010 at 10:16 AM, Robby Findler <[email protected]> wrote: > Probably bestto switch this to the users mailing list. > > The short answer: check out places and futures in the docs. Those are > both pretty new but that's what we have so far. > > Robby > > On Sunday, October 3, 2010, Christopher Charles Lanz <[email protected]> > wrote: >> Greets, >> >> I was at last Summer's Teach Scheme Not workshop, and was so impressed >> that I want to switch over to some form of Racket/Scheme/Lisp for my >> research. >> >> But I can't unless there's some form of Racket or Scheme that can be used >> with an MPI cluster. (Some of the cluster nodes will also have Nvidia GPU >> co-processors, so somehow I have to be able to link to CUDA, but that can >> come later.) >> >> Whatever shall I do?! >whine< >snivel< >> >> I'd prefer Racket, 'cuz we might start using it for PROGRAMMING 101. >> >> >> Chris Lanz >> 315 268 1547 >> [email protected] >> >> _________________________________________________ >> PLT Educators mailing list >> [email protected] >> http://list.cs.brown.edu/mailman/listinfo/plt-edu >> > _________________________________________________ > PLT Educators mailing list > [email protected] > http://list.cs.brown.edu/mailman/listinfo/plt-edu > -- Jay McCarthy <[email protected]> Assistant Professor / Brigham Young University http://teammccarthy.org/jay "The glory of God is Intelligence" - D&C 93 _________________________________________________ For list-related administrative tasks: http://lists.racket-lang.org/listinfo/users
