Simulation and analysis of a neuromorphic architecture for in-memory computing

Abstract

Memristor based systems are proposed to break away from the classical von Neumann architecture and emulate neuromorphic behavior, aiming to address the energy consumption challenges posed by current computing systems. Modeling a memristor based on experimental measurements enables the simulation and analysis of NOT and NOR gates based on NMOS-like RRAM architecture. This study delves into said device’s properties, possible systems implemented with it including in-memory computation, and analyses of energy efficiency