( on using 1024 cpus)
>Binding that many CPUs together is hard! You tell me how you plan to do
>it, and _then_ I'll tell you whether clustering is faster. :)
I'd do it the same way that KSR did, except that I'd use off the shelf
processors. Cornell, I think, had a 1024 node KSR-1 and maybe even a
1024 KSR-2. They never did get all the kinks out of their cache
coherency protocol, but I think that between them and Alewife, it
could probably be done.
Next question, however: why ? Problems that benefit from this level of
parallelism using general purpose microprocessors (as distinct from
the PDP-neural-network kinds of parallelism) are typically well served
by clustering.
This "why" question, along with the fact that UP speeds keep making
various MP configurations redundant a couple of years after their
owners spent big money on them, have been the bugbear of most
large-scale MP designs.
--p (still angry that UWashington could spend $1M on a KSR and then
throw it away a year later)
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/