Evaluation of IGZO Synapses for Neuromorphic Systems

Abstract

Artificial intelligences are the most important technology in future societies. However, they are realized as complicated software executed on high-spec hardware, and it is issues that the size is large, and the power consumption is huge. Therefore, it is difficult to incorporate artificial intelligences into IoT equipment. The neuromorphic architecture is a system that imitates the human brain computing system and realizes human brain functions. We fabricated the neuromorphic device using an LSI and IGZO thin films. We evaluated the synaptic characteristics of IGZO thin films and confirmed the basic operation of the synapse device.