Artificial Synapses: Low‐Power, Electrochemically Tunable Graphene Synapses for Neuromorphic Computing (Adv. Mater. 36/2018)

Abstract

By reversibly intercalating ions between the layers of two-dimensional graphene, Feng Xiong and co-workers at Pitt develop a novel artificial synapse for neuromorphic computing, as described in article number 1802353. With over 250 tunable analog states, good energy efficiency, and promising scalability, these electrochemical synapses can lead to the hardware implementation of neural networks and hence the prevalent use of artificial intelligence.