Photonic co-processors in HPC: using LightOn OPUs for Randomized Numerical Linear Algebra

TL;DR

This paper demonstrates how LightOn Optical Processing Units can accelerate randomized linear algebra algorithms in HPC, reducing computational bottlenecks with minimal precision loss.

Contribution

It introduces the use of LightOn OPUs for fast random projections, significantly improving the efficiency of RandNLA methods in high-performance computing.

Findings

Random projections are accelerated using LightOn OPUs.

Significant reduction in computational time for RandNLA algorithms.

Negligible precision loss in accelerated algorithms.

Abstract

Randomized Numerical Linear Algebra (RandNLA) is a powerful class of methods, widely used in High Performance Computing (HPC). RandNLA provides approximate solutions to linear algebra functions applied to large signals, at reduced computational costs. However, the randomization step for dimensionality reduction may itself become the computational bottleneck on traditional hardware. Leveraging near constant-time linear random projections delivered by LightOn Optical Processing Units we show that randomization can be significantly accelerated, at negligible precision loss, in a wide range of important RandNLA algorithms, such as RandSVD or trace estimators.