Scientists have created memristors that can simultaneously store and process information
Engineers have developed graphene-based memristors that act as synapses in the brain, capable of not only storing information, but also processing it.
Modern computers can perform an enormous amount of computation per unit of time. The problem is that they store data in one place and process it in another. This architecture makes them cumbersome, and the need to constantly move information requires significant energy consumption and limits performance. In addition, unmanned vehicles, artificial intelligence and big data technologies need to develop more efficient types of computing.
Best CryptoCurrency Exchange Sites To Buy/Sell All CryptoCurrency
Scientists at Penn State University have developed artificial neural networks that work like synapses in the brain that can be reconfigured by applying a short-term electric field to a graphene sheet. The team’s research showed that the memristors they created support at least 16 possible memory states.
The technology allows the development of systems that can simultaneously store and process information in one place, which will contribute to the development of neuromorphic computing.
We also previously reported on the creation of memristors from protein nanowires, which are able to branch out and form new connections, mimicking the learning process of the brain.