MIT Engineers Beat the Complexity of Deep Parallel Processing

lotuseclat79 · Feb 23, 2015

MIT Engineers Beat the Complexity of Deep Parallel Processing.

Engineers at MIT have developed an algorithm that promises to boost the speeds of multicore systems by nearly 50 percent—while upping their overall power efficiency by some 30 percent in the process. It doesn't entirely nuke the lingering parallel processing memory bottleneck known as the cache hierarchy, but the new scheme is poised to at least widen it considerably. Computers can get faster after all. Phew.
Click to expand...

-- Tom

Log in or Sign up

MIT Engineers Beat the Complexity of Deep Parallel Processing

lotuseclat79 Registered Member

Log in or Sign up

MIT Engineers Beat the Complexity of Deep Parallel Processing

lotuseclat79 Registered Member

Useful Searches