LightOn Optical Processing Unit: Scaling-up AI and HPC with a Non von Neumann co-processor

TL;DR

The paper presents LightOn's Optical Processing Unit, a photonic AI accelerator chip capable of 1500 TeraOPS, integrating optics with software for scalable non von Neumann computations in AI and HPC.

Contribution

Introduction of the first market-ready photonic AI accelerator chip combining optics and software API for scalable non von Neumann computing.

Findings

Achieves 1500 TeraOPS performance

Supports seamless integration with Python pipelines

Enables hybrid CPU/GPU and optical architectures

Abstract

We introduce LightOn's Optical Processing Unit (OPU), the first photonic AI accelerator chip available on the market for at-scale Non von Neumann computations, reaching 1500 TeraOPS. It relies on a combination of free-space optics with off-the-shelf components, together with a software API allowing a seamless integration within Python-based processing pipelines. We discuss a variety of use cases and hybrid network architectures, with the OPU used in combination of CPU/GPU, and draw a pathway towards "optical advantage".