Reconfigurable Spin Switch
Researchers at Purdue University have developed technology for a hardware neural network implementation using a spin switch (SS) as its basic building block. This allows neural networks to be constructed with purely passive interconnections without intervening clocks or amplifiers. The weights for the neural network are conveniently adjusted through analog voltages that can be stored in a non-volatile manner in an underlying CMOS layer using a floating gate low drop out voltage regulator. The operation of a multilayer SS neural network designed for character recognition is demonstrated using a standard simulation model based on coupled Lanadau-Lifshitz-Gilbert equations, one for each magnet in the network.
Requires less power than current methods Unit size is relatively small
Computer networks
Supriyo DattaPurdue Electrical and Computer Engineering
United States
None
美国
