Communications of the ACM
New 'artificial Synapses' Pave Way For Brain-Like Computers
Researchers at University of Massachusetts Amherst have developed memristors that mimic the behavior of calcium ions in the synapses between neurons in the human brain.
Credit: Billion Photos/Shutterstock.com
The memristor is an electrical component that mimics the behavior of calcium ions in the synapses between neurons, and may lead to significant advances in brain-inspired computers.
The reproduction of synaptic plasticity is essential for creating computers that can operate like the brain, which is far more efficient than traditional electronics, according to University of Massachusetts Amherst professor Joshua Yang. He and his team developed a new diffusive memristor consisting of silver nanoparticles embedded in a silicon oxynitride film between two electrodes. When a voltage pulse is applied, nanoparticles diffuse through the film to form a filament that carries the current from one electrode to the other. The diffusion memristor also can be combined with a drift memristor, which relies on electrical fields and is optimized for memory applications.
"The diffusion memristor is helping the drift-type memristor behave similarly to a real synapse," Yang says. "Combining the two leads us to a natural demonstration of [spike-timing-dependent plasticity], which is a very important long-term plasticity learning rule."
Researchers use a fabrication process similar to those developed by computer memory companies to scale up production of the memristors, and although the device could hypothetically be made smaller than a human synapse, multi-unit systems will require larger devices.
From LiveScience
View Full Article
Abstracts Copyright © 2016 Information Inc., Bethesda, Maryland, USA
No entries found