Simulations of the Effects of Charge Traps in the Ferroelectric Hafnium-Oxide Gate Stack of Ferroelectric Field Effect Transistors (FeFETs) Under Pulsed Operation for Spiking Neural Networks

Author(s):

Pankaj B. Shah

Abstract:

Developing ferroelectric field effect transistors (FeFETs) for compute-in-memory and neuromorphic applications is a growing field of research. Technology computer-aided design drift-diffusion simulation results obtained at ARL for ferroelectric transistors and capacitors are presented in this report. Three leading commercially available software were considered for simulations to understand the status of commercially available software and determine where gaps exist. FeFET simulation results demonstrate how charge trapping would affect FeFET performance as a neuron or synaptic element in a spiking neural network.