Re: processors of the future: super-computer-on-a-chip?

Joe Seigh wrote:
Bill Todd wrote:
Chris Thomasson wrote:



" How many cores do you think a chip could have, lets say 10 or 20+ years from now? "


Well, only partially playing devil's advocate here, I'll counter with the question:

"Who cares?"

...


IBM (nobody's fool these past few years) certainly doesn't seem to be relegating single-threaded performance to the back seat of its POWER architecture: instead, after pioneering dual-core products 5 years ago it has been steadily improving their single-threaded performance (with only one additional nod to multi-threading in its SMT facilities which, perhaps not coincidentally, likely offer some single-thread performance advantages as well to threads with sufficient ILP). And only Sun has introduced a product which has *conspicuously* traded off single-thread for multi-core performance (a product which Sun itself agrees is suitable only for certain niches rather than being really general-purpose).

Sure, there will be a few applications that could make really good use of huge numbers of slower cores, but will they fund the associated development sufficiently to overcome the resources available to develop commodity products (always the lament of anything special-purpose these days)? And for that matter aren't there any better uses we can come up with for chip area?

Maybe in a decade or more some real application shift in this direction will at least *start* to occur, but I don't expect it nearly as soon as it becomes *possible* to build such cores, so I don't find it particularly interesting save as an intellectual exercise for now. Until we can at least come up with ways for software to use such hypothetical hardware to solve some respectable range of problems far better than they can be solved using obvious extensions to today's mechanisms, it seems uncomfortably close to debating how many angels can dance on the head of a pin.


Intel at least wants to know how fast those angels can dance.
http://www.theregister.com/2006/08/22/intel_suite_help/
And maybe a few new dances as well.

Since I'd interpret most of that article as applying to chips with under 10 or so cores (in fact, the only example given is a 4-core chip), I'm not sure how you came to that conclusion. True, it states (though rather nebulously)

[quote]

"From everything we can see, there are certainly interesting things to do with tens of cores and hundreds of threads," Rattner said. "That's where we are targeted with our research."

[end quote]

but then immediately continues with

[quote]

Rattner urged chip makers not to get caught in a "core war" where they make myriad multi-core chips just because they can. Software and tools vendors have a lot of catching up to do before such chips will be useful in broad terms, Ratter said.

[end quote]

which somewhat resembles the observation I made myself.

....

It's probably as short sighted to say that we will never need more
then X cores as it was to say the world will never need more than
5 computers.

That could conceivably be why I suggested nothing of the kind, but rather that *some* broad, significant use should at least be envisionable for such products before the industry commits whole-heartedly to embracing them.

- bill
.