High Performance 2D Perovskite/Graphene Optical Synapses as Artificial Eyes

Conventional von Neumann architectures feature large power consumptions due to memory wall. Partial distributed architecture using synapses and neurons can reduce the power. However, there is still data bus between image sensor and synapses/neurons, which indicates plenty room to further lower the power consumptions. Here, a novel concept of all distributed architecture using optical synapse has been proposed. An ultrasensitive artificial optical synapse based on a graphene/2D perovskite heterostructure shows very high photo-responsivity up to 730 A/W and high stability up to 74 days. Moreover, our optical synapses has unique reconfigurable light-evoked excitatory/inhibitory functions, which is the key to enable image recognition. The demonstration of an optical synapse array for direct pattern recognition shows an accuracy as high as 80%. Our results shed light on new types of neuromorphic vision applications, such as artificial eyes.