On Thursday 18 December 2008, Dan wrote: > I have been using Sage for a fairly intense computational biology > project. I have been very pleased with the software, but when it was > finely time to run my computations I learned that my hardware was > grossly inadequate. I have since gained access to a computer cluster > (256 G5 xserves) that seems adequate to handle my need, but the > cluster manager asked about the ability of Sage to be distributed > across processors. I am writing to possibly get answers for him. The > manager and I reviewed the information about DSage, but questions > remain. Specifically, is Sage (or the modules I am using within Sage: > R, Numpy, SQLalchemy) capable of being "fully distributed", or is the > grid computing (i.e. "coarse distribution"?) the only option.
I am not quite sure what you mean by 'fully distributed' but if you expect it to just work (TM) on 256 nodes then I guess you're out of luck. Numpy should make use of ATLAS which can be tuned to use all available cores in a multicore, shared memory setting. SQLalchemy probably depends on the database you're interfacing with. R: no clue. I suspect your best shot is more or less embarrassingly parallel coarse grained stuff. However, without a specific example it is hard to say. Hope that helps, Martin -- name: Martin Albrecht _pgp: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x8EF0DC99 _www: http://www.informatik.uni-bremen.de/~malb _jab: martinralbre...@jabber.ccc.de --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
