***intel's new neuromorphic self learning chip ''Codenamed Lowie***
If Intel's new Artificial Intelligence project goes to the market, it can make the future different. They announce that this new neurorphic chip, Codenamed Lohehi has been designed for the complexity of Ai's complex research. Dr. Michael Michaelby of Intel claims that Loyhi is not usually trained in the stream, which is trained in new ways through a computing system called asynchronous spikeing.
According to Intel
The neural network of brain controls the data through pulse and spike, based on the synchronous energy interconnection, based primarily on Spike timing. The intelligence is triggered by collaborative and competitive interconnection in the brain's neural network.
Where neural spike models are very useful to work on biological cells, they apply this model to solve real-time computing and engineering problems. Which is exceptionally effective for cutting-edge applications, but the real-time application of this power has not yet been fully opened!
Intel claims that MNIS is more than 100 times faster than normal neural spike net solution to solve digit Recognition problem. Although it does not say how the normal neural network is formed. Intel also claims that it is more effective in the field of coiled neural networks and in complex research areas.
Cheeppe's neurons can maintain contact with thousands of other neurons. Which exists as a neuromorphic core learning engine. It has 130000 neurons, with 130 million synapses. Which is significantly less than IBM's "Trunthth", IBM created it through 4096 cores with one million programmable neurons and 256 million synopses three years ago.