Researchers at MIT have developed a new CPU that mimics the manner in which the brain's synapse adapts to new information. This phenomenon is known as plasticity and is believed to underlie many brain functions, including learning and memory.
In order to design such a chip, the researchers have to come up with a computer processor that is analog in nature because a traditional digital computing processor is unable to mimic plasticity. In order to simulate the activity of a single brain synapse, the processor requires about 400 transistors. Each synapse allows information to flow from one brain neuron to the other. A adult human brain is estimated to have 100 trillion synapses (that is 10 to the power of eighteen); hence, the CPU being developed would need to have 40 trilliard transistors or 40 thousand trillion transistors. This kind of computing horsepower would translate to using over 18 thousand million Intel Core i7-3960X CPUs!
The MIT researchers plan to implement their chip on a smaller scale to build systems to model specific neural functions, such as the visual processing system. This technology is still in its infancy, but the MIT researchers believe that this chip will help neuroscientists learn much more about the brain functions, and could also be used in neural prosthetic devices such as artificial retinas.
Source: MIT News