Brain-Like Device Simulating Dendrites Perception Process and Optical Induced Excitatory Postsynaptic Current

Abstract

The traditional von Neumann computing architecture leads to hidden costs in terms of computing resources and energy demand, and the brain-inspired neuromorphic artificial intelligence architecture is receiving increasing attention as one of the main competitors in improving computing power. Based on the issues above, a brain-like device based on Mg_0.1Zn_0.9O thin film is fabricated. This device can simulate various synaptic behaviors, including typical long/short-term plasticity. In neuromorphic computing, an artificial neural network is built to achieve handwritten digit recognition. The accuracy of recognition was improved through the nonlinearity modulation of synaptic weights (conductance). The brain-like device simulated dendrite sensing processes and exhibits four behavioral outcomes. Excitatory postsynaptic current (EPSC) is induced by 365 nm ultraviolet stimulation with an intensity of 23.5 mW/cm² on the brain-like device. The memory effect of EPSC is modulated through changing the duration of light, which is similar to the learning process of the human brain and shows potential in optical neuromorphic devices.