An artificial analog synapse that can run 1 million times faster than the biological ones in the human brain has been developed by a team of researchers at the Massachusetts Institute of Technology (MIT). The newly developed programmable resistors can be used to make up analog neural networks in artificial intelligence systems and can process information around a million times faster than brain synapses that link neurons together.
The latest device is cheaper to build, and it is claimed to be more energy efficient. The researchers used programmable resistors in repeating arrays to create a complex layered network of artificial neurons and synapses, which executed computations just like a digital neural network. The network can be used to achieve complex AI tasks like image recognition and neural language processing.