A Highly Efficient and Scalable Model for Crossbar Arrays with Nonlinear Selectors

Abstract

A scalable crossbar array model is proposed to solve arrays with various sizes and device characteristics. Computation is accelerated >1000× by array reduction, enabling array scaling analysis and extensive design space exploration. Computation errors can be reduced by extrapolation from calculations with different array reduction ratios. The impact of selectors on array performance is analyzed with this model, which can be applied to nonlinear, threshold switch, and rectifying selectors.