Character inference learning for stacked neuromorphic devices using IGZO thin films

We conducted research and development of a hardware neural network by using oxide semiconductors of a-In-Ga-Zn-O (IGZO). We have made a 3 layers cross-point type device that uses electrical characteristics of IGZO as a synapse and a metal electrode as an axon of a neuron. It is possible to enable parallel computing, high operating speed, low power consumption, high integration, and robustness, that were not possible with software. By designing a learning method using a three-layer device created with the goal of high integration and simulating it, its practicality was demonstrated.