Investigation of Analog Memristor Characteristics for Hardware Synaptic Weight in Multilayer Neural Network

Abstract

Analog Memristor Characteristics

The systematic design of memristor-based neural network is provided by analog conductance state parameters to accurately emulate the software-based high-resolution weight at discrete device level. The requirement of discrete analog conductance of memristor device is measured as ≈50 states with nonlinearity value of ≈0.142 within the deviation range of 5% for inference accuracy of ≈84.36% and loss value of ≈0.168. Further details can be found in article number 2400710 by Jingon Jang, Yoonseok Song, and Sungjun Park.