Proposal and Experimental Demonstration of Reservoir Computing using Hf0.5Zr0.5O2/Si FeFETs for Neuromorphic Applications

Abstract

We propose a new AI calculation scheme by reservoir computing utilizing the memory effect and nonlinearity of ferroelectric gate MOSFETs (FeFETs) for neuromorphic applications. The task operations of time-series data are experimentally demonstrated by taking time responses of the drain current for gate voltage input as the virtual nodes. A high ability to classify input data is experimentally verified.