On Memory System Design for Stochastic Computing

TL;DR

This paper introduces StochMem, an analog memory system designed specifically for stochastic computing, significantly reducing energy and area overheads while maintaining high computational accuracy.

Contribution

It presents the first dedicated memory system design for stochastic computing, integrating analog memory to optimize energy, area, and accuracy trade-offs.

Findings

Reduces energy consumption by up to 52.8%.

Lowers area overhead by up to 93.7%.

Maintains accuracy with at most 0.7% loss.

Abstract

Growing uncertainty in design parameters (and therefore, in design functionality) renders stochastic computing particularly promising, which represents and processes data as quantized probabilities. However, due to the difference in data representation, integrating conventional memory (designed and optimized for non-stochastic computing) in stochastic computing systems inevitably incurs a significant data conversion overhead. Barely any stochastic computing proposal to-date covers the memory impact. In this paper, as the first study of its kind to the best of our knowledge, we rethink the memory system design for stochastic computing. The result is a seamless stochastic system, StochMem, which features analog memory to trade the energy and area overhead of data conversion for computation accuracy. In this manner StochMem can reduce the energy (area) overhead by up-to 52.8% (93.7%) at the cost of at most 0.7% loss in computation accuracy.