<html><body>
<DIV><BR>-------------- Original message -------------- <BR>From: Greg Lindahl <lindahl@pbm.com> <BR><BR>> On Thu, Aug 23, 2007 at 09:09:57AM -0400, Douglas Eadline wrote: <BR>> <BR>> > Naturally, if you have four processes running <BR>> > it is best if each one gets its own woodcrest. To the OS <BR>> > the all look the same. Other than Intel MPI, I don't <BR>> > know of any other MPI that attempts to optimize this. <BR>> <BR>> InfiniPath MPI has always optimized this. Of course, there's no way it <BR>> can read your mind and know if you are going to run a second 4-core <BR>> job on the same node, so there is no perfect default. But it has <BR>> switches to give you either behavior, tightly packed or spread out. </DIV>
<DIV> </DIV>
<DIV>An additional example is Liquid Computing's implementation of MPI</DIV>
<DIV>for their 4x2 way mother boards. You have complete control of </DIV>
<DIV>placement (not withstanding Greg's mind reading point).</DIV>
<DIV> </DIV>
<DIV>rbw<BR>-- <BR><BR>"Making predictions is hard, especially about the future." <BR><BR>Niels Bohr <BR><BR>-- <BR><BR>Richard Walsh <BR>Thrashing River Consulting-- <BR>5605 Alameda St. <BR>Shoreview, MN 55126 <BR><BR>Phone #: 612-382-4620
<DIV> </DIV></DIV>
!DSPAM:46cdea8a290341523621093!
</body></html>