Unified Memcapacitor-Memristor Memory for Synaptic Weights and Neuron Temporal Dynamics

TL;DR

This paper introduces a novel memory device combining memristive and memcapacitive properties, enabling advanced control of neural network dynamics for more efficient neuromorphic computing.

Contribution

The work presents a fabricated memory stack with dual memristive and memcapacitive behavior, and a circuit design for improved neural network control.

Findings

Hardware-aware simulations show promising neuromorphic processing capabilities.

Experimental characterization confirms dual memristive-memcapacitive functionality.

The memory device enables simultaneous control of spatial and temporal neural dynamics.

Abstract

We present a fabricated and experimentally characterized memory stack that unifies memristive and memcapacitive behavior. Exploiting this dual functionality, we design a circuit enabling simultaneous control of spatial and temporal dynamics in recurrent spiking neural networks (RSNNs). Hardware-aware simulations highlight its promise for efficient neuromorphic processing.