PCM-trace: Scalable Synaptic Eligibility Traces with Resistivity Drift of Phase-Change Materials

TL;DR

This paper introduces PCM-trace, a novel neuromorphic building block using phase-change materials to implement long-lasting eligibility traces, enabling scalable, efficient on-chip learning in spiking neural networks with experimental validation.

Contribution

It presents a new PCM-trace device leveraging resistivity drift for eligibility traces, significantly improving area efficiency and supporting three-factor learning rules in neuromorphic hardware.

Findings

Achieves over 10x area reduction compared to existing solutions.

Demonstrates experimentally the feasibility of resistivity drift for long-lasting eligibility traces.

Supports a techno-logically plausible learning algorithm with device measurements.

Abstract

Dedicated hardware implementations of spiking neural networks that combine the advantages of mixed-signal neuromorphic circuits with those of emerging memory technologies have the potential of enabling ultra-low power pervasive sensory processing. To endow these systems with additional flexibility and the ability to learn to solve specific tasks, it is important to develop appropriate on-chip learning mechanisms.Recently, a new class of three-factor spike-based learning rules have been proposed that can solve the temporal credit assignment problem and approximate the error back-propagation algorithm on complex tasks. However, the efficient implementation of these rules on hybrid CMOS/memristive architectures is still an open challenge. Here we present a new neuromorphic building block,called PCM-trace, which exploits the drift behavior of phase-change materials to implement long lasting eligibility traces, a critical ingredient of three-factor learning rules. We demonstrate how the proposed approach improves the area efficiency by >10X compared to existing solutions and demonstrates a techno-logically plausible learning algorithm supported by experimental data from device measurements