Two-Dimensional Materials for In-Sensor and Near-Sensor Computing

In-sensor computing requires devices that can perform sensing along with storing the sensed information. New devices, such as optoelectronic synapses, can sense light and store the information as tunable conductance states. Two-dimensional (2D) materials enable such devices because of their superior optoelectronic properties. These devices are designed to optimize light sensitivity and electronic properties, allowing them to be integrated to perform vector matrix multiplication at the sensor node. Memristive devices with 2D materials are also used to form synaptic crossbar arrays, which can be heterogeneously integrated with CMOS devices. The back-end-of-line compatibility of 2D materials allows for the possibility of a 3D integrated neuromorphic system for edge AI.