When Pipelined In-Memory Accelerators Meet Spiking Direct Feedback Alignment: A Co-Design for Neuromorphic Edge Computing

TL;DR

This paper presents a co-designed hardware-software approach for efficient training of Spiking Neural Networks on edge devices, combining a novel training algorithm with an optimized in-memory computing architecture.

Contribution

It introduces Spiking Direct Feedback Alignment (SDFA), a hardware-friendly training algorithm, and a pipelined IMC architecture, PipeSDFA, for accelerated SNN training on resource-constrained devices.

Findings

Achieves less than 2% accuracy loss on five datasets.

Reduces training time by up to 10.5 times.

Lowers energy consumption by up to 2.1 times.

Abstract

Spiking Neural Networks (SNNs) are increasingly favored for deployment on resource-constrained edge devices due to their energy-efficient and event-driven processing capabilities. However, training SNNs remains challenging because of the computational intensity of traditional backpropagation algorithms adapted for spike-based systems. In this paper, we propose a novel software-hardware co-design that introduces a hardware-friendly training algorithm, Spiking Direct Feedback Alignment (SDFA) and implement it on a Resistive Random Access Memory (RRAM)-based In-Memory Computing (IMC) architecture, referred to as PipeSDFA, to accelerate SNN training. Software-wise, the computational complexity of SNN training is reduced by the SDFA through the elimination of sequential error propagation. Hardware-wise, a three-level pipelined dataflow is designed based on IMC architecture to parallelize the training process. Experimental results demonstrate that the PipeSDFA training accelerator incurs less than 2% accuracy loss on five datasets compared to baselines, while achieving 1.1X~10.5X and 1.37X~2.1X reductions in training time and energy consumption, respectively compared to PipeLayer.