Think of all the things you could do with a 100-core processor, or even a 1,000-core processor. Climb the ranks of Maximum PC's Folding@Home team! Encode videos like a boss! Run Crysis! Ah, if only it was as easy as piling on more cores for exponential performance gains. There's more to it than that, and a couple of researchers from North Carolina University say they've developed a pair of techniques that will help maximize the performance of multi-core processors by allowing them to retrieve data more efficiently.
As the researchers explain it, the new techniques allow multi-core chips to do a better job allocating bandwidth and "prefetching" data, according to a press release issued by NCSU.
"The first technique relies on criteria we developed to determine how much bandwidth should be allotted to each core on a chip," says Dr. Yan Solihin, associate professor of electrical and computer engineering at NC State and co-author of a paper describing the research. "By better distributing the bandwidth to the appropriate cores, the criteria are able to maximize system performance.
"The second technique relies on a set of criteria we developed for determining when prefetching will boost performance and should be utilized, as well as when prefetching would slow things down and should be avoided."
By using both techniques, the researchers claim they were able to boost multi-core chip performance by 40 percent compared to multi-core processors that don't prefetch data, and 10 percent compared to multi-core chips that always prefetch data.
For all the geeky details, you can read the 12-page paper here .