Multidimensional photonic computing

TL;DR

This paper discusses the development of multidimensional photonic computing platforms that leverage the unique properties of photons to significantly enhance computational throughput, bandwidth, and energy efficiency beyond traditional electronic systems.

Contribution

It introduces recent advancements in multidimensional photonic computing architectures, highlighting their potential to overcome limitations of electronic computing for AI workloads.

Findings

Photonic systems enable processing in multiple orthogonal dimensions.

Multidimensional photonic computing promises low-latency, high-bandwidth data processing.

Potential for reduced energy consumption compared to electronic counterparts.

Abstract

The rapidly increasing demands for computational throughput, bandwidth, and memory capacity fueled by breakthroughs in machine learning pose substantial challenges for conventional electronic computing platforms. For digital scaling to keep pace with the accelerating growth of artificial intelligence (AI) models beyond the trajectory of Moores law, computational power has to double roughly every three months. Historically, advancing compute performance relied on spatial scaling to increase the transistor count on a given chip area and, more recently, the development of parallel and multi-core architectures. Exponential scaling on trajectories much steeper than what can be achieved by such conventional strategies, and in line with the demands of AI, can be achieved with computing platforms that process data using multiple, orthogonal dimensions available to photons. Here we elucidate pivotal developments in the realization of multidimensional computing platforms based on photonic systems. Moving to such architectures holds enormous promise for low-latency, high-bandwidth information processing at reduced energy consumption.