A Scalable Approach to Performing Multiplication and Matrix Dot-Products in Unary

TL;DR

This paper introduces a scalable deterministic method for performing multiplication and matrix dot-products in unary, reducing latency and area compared to traditional stochastic computing, with applications in machine learning.

Contribution

It presents a novel approximation technique that maintains low latency and scales efficiently, improving upon existing stochastic computing methods for matrix operations.

Findings

Outperforms traditional stochastic designs in matrix multiplication

Area scales with the square root of the precision parameter

Latency is significantly reduced compared to previous deterministic approaches

Abstract

Stochastic computing is a paradigm in which logical operations are performed on randomly generated bit streams. Complex arithmetic operations can be executed by simple logic circuits, resulting in a much smaller area footprint compared to conventional binary counterparts. However, the random or pseudorandom sources required for generating the bit streams are costly in terms of area and offset the advantages. Additionally, due to the inherent randomness, the computation lacks precision, limiting the applicability of this paradigm. Importantly, achieving reasonable accuracy in stochastic computing involves high latency. Recently, deterministic approaches to stochastic computing have been proposed, demonstrating that randomness is not a requirement. By structuring the computation deterministically, exact results can be obtained, and the latency greatly reduced. The bit stream generated adheres to a "unary" encoding, retaining the non-positional nature of the bits while discarding the random bit generation of traditional stochastic computing. This deterministic approach overcomes many drawbacks of stochastic computing, although the latency increases quadratically with each level of logic, becoming unmanageable beyond a few levels. In this paper, we present a method for approximating the results of the deterministic method while maintaining low latency at each level. This improvement comes at the cost of additional logic, but we demonstrate that the increase in area scales with the square root of n, where n represents the equivalent number of binary bits of precision. Our new approach is general, efficient, composable, and applicable to all arithmetic operations performed with stochastic logic. We show that this approach outperforms other stochastic designs for matrix multiplication (dot-product), which is an integral step in nearly all machine learning algorithms.