Saturday, April 26, 2008

Utility of Multicore Chips?

I have to admit that I'm a skeptic of multicore chips. Even though they're part of many CPUs shipping today, it's just not clear to me what problem they solve. We've already had marginal returns on CPU performance, in terms of human productivity. The bottleneck just isn't the microprocessor anymore.

Don Knuth also sees some challenges for multicore:

Let me put it this way: During the past 50 years, I’ve written well over a thousand programs, many of which have substantial size. I can’t think of even five of those programs that would have been enhanced noticeably by parallelism or multithreading. Surely, for example, multiple processors are no help to TeX.


On the other hand, multicore might represent an opportunity. One of the trends in research this past decade has been novel ways of "wasting" CPU to enable other desirable properties, such as security and usability. All we need now is a clearer path for making this happen.

No comments: