Hybrid FeRAM/RRAM Synaptic Circuit Enabling On-Chip Inference and Learning at the Edge

Abstract

This article presents an experimental demonstration of a hybrid FeRAM/RRAM synapse circuit implemented at the 130nm node. The circuit incorporates Metal-Ferroelectric-Metal (MFM) stacks, which exhibit FeRAM capacitor behavior when no filament formation occurs and function as RRAM after undergoing a forming operation. By leveraging the unique advantages of FeRAMs, such as ultra-low power consumption, in combination with the nondisruptive (infinite) reading capability of RRAMs, this circuit enables efficient on-chip inference and learning at the Edge.