Eyeriss is MIT’s 168-Core Chip That Brinngs Powerful AI to Mobiles
We have known different services to offer features like facial recognition, object and sound recognition but for that, all those services have to rely on the cloud where they tap in every time to check with the memory. However, Eyeriss is a project by MIT that will be able to do that on its own and for mobile devices – but that is just the beginning of its capabilities.
This has the potential of taking smartphone based artificial intelligence to a whole new level, a product where the need for a cellular data will be minimal even in the case of such complex tasks because each of its cores has its own memory bank.
We are talking about a 168-core chip that looks to empower robots, self-driving cars, drones and smartphones with neural networks.
In comparison with the high-end mobile GPUs we know now, this will be 10 times more powerful yet energy-friendly allowing them to “run powerful artificial-intelligence algorithms locally, rather than uploading data to the Internet for processing.”
Eyeriss was presented at the International Solid State Circuits Conference in San Francisco this week by a team of MIT researchers where the deep learning power of the chip was shown off.
With powerful artificial-intelligence algorithms on board, networked devices could make important decisions locally, entrusting only their conclusions, rather than raw personal data, to the Internet. And, of course, onboard neural networks would be useful to battery-powered autonomous robots.
So far we are not sure if and when Eyeriss chips will make it to consumer products, but we will surely love too see them in action.