Compute-in-Memory Implementation of State Space Models for Event Sequence Processing

TL;DR

This paper presents a novel compute-in-memory hardware implementation of state space models that enables real-time, energy-efficient, event-driven processing for long sequence tasks, outperforming traditional methods.

Contribution

It introduces a hardware-friendly re-parameterization of SSMs and a co-design approach leveraging device dynamics for efficient implementation in CIM systems.

Findings

Achieves high accuracy in event-based vision and audio tasks.

Demonstrates significant energy efficiency improvements.

Supports fully asynchronous processing.

Abstract

State space models (SSMs) have recently emerged as a powerful framework for long sequence processing, outperforming traditional methods on diverse benchmarks. Fundamentally, SSMs can generalize both recurrent and convolutional networks and have been shown to even capture key functions of biological systems. Here we report an approach to implement SSMs in energy-efficient compute-in-memory (CIM) hardware to achieve real-time, event-driven processing. Our work re-parameterizes the model to function with real-valued coefficients and shared decay constants, reducing the complexity of model mapping onto practical hardware systems. By leveraging device dynamics and diagonalized state transition parameters, the state evolution can be natively implemented in crossbar-based CIM systems combined with memristors exhibiting short-term memory effects. Through this algorithm and hardware co-design, we show the proposed system offers both high accuracy and high energy efficiency while supporting fully asynchronous processing for event-based vision and audio tasks.