Last week a IEEE report claimed multi-core processors are bad news for supercomputers because of bandwidth problems but both Intel and AMD denied this claim, stating they can compete in the high performance computing (HPC) arena.
But in some specific circumstances, multi-core isn't very helpful. Sandia engineers found is as more cores are added, the more performance degrades because the memory can't keep up with the cores. Instead of one CPU talking to the memory, 4 CPUs are now talking to the memory. That effectively means one-quarter of the memory performance.
Intel said it is addressing the issue in its TeraScale 80-core chip. "Intel's work on stacking memory could be key to resolving long term multicore memory bottlenecks, but this was not discussed in the article," said an Intel spokesperson in a statement e-mailed to InternetNews.com.
"We've been talking in public about the need to integrate memory closer to processors for more than two years now; showing directionally what will need to happen. We are confident that the industry will work around the memory bandwidth issue, currently there are very fast memory bandwidth subsystems, but reasonable cost as well as performance needs to be achieved," the statement went on to say.