News for hackers. Stuff that matters.
Re:Microsoft Supercharges Bing Search With Programmable Chips by colanderman on Monday June 16, 2014 @11:17PM (#7902708)
I think what's more likely is that processing will move into the memory itself. We already see this somewhat with manycore shared-cache architectures like Tilera, which effectively provide one core per ¼ MiB of L3 cache. (Though unfortunately L3 cache latency is only ~twice as fast as RAM, but at least the bandwidth is much better.)

There are no comments on this comment yet.