[Beowulf] Inside Tsubame - the Nvidia GPU supercomputer - OpenCL

[Daniel Kidger]

Tsubame isn't just about delivering flops to production codes.
It is trying to spearhead coprocessing.
The people there have been porting codes to ClearSpeed / GPUs for a
while - and hence been publishing their experiences.

Daniel


On Fri, 2008-12-12 at 12:01 +0000, Loic Tortay wrote:
> Florent Calvayrac-Castaing wrote:
> [...]
> >
> > Interesting.
> >
> > I understand why, when I submitted a joint exploratory project
> > about GPU computing two years ago with a Japanese
> > colleague we were ranked first in Japan and last in France ; the
> > idea seems more popular in Japan if they can fork  millions
> > on an architecture it is not very quick to program for (at least
> > maybe not as fast as Moore's law is increasing power).
> >
> They may be willing to spend millions because they already have programs
> able to use the GPUs.
> 
> If I'm not mistaken, the "Tsubame" cluster was initially using
> Clearspeed accelerators (in Sun X4600 "fat" nodes).
> 
> Therefore, they probably have appropriate programs that need little
> adaptation (or less than many) to work on the GPUs.
> 
> 
> Loïc.
> 
> _______________________________________________
> Beowulf mailing list, Beowulf at beowulf.org
> To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf