Towards Topology-Aware Very Large-Scale Photonic AI Accelerators

TL;DR

This paper introduces a scalable, topology-aware photonic AI accelerator architecture that addresses system-level scalability challenges and demonstrates significant performance and efficiency improvements through topology optimization.

Contribution

It presents a modular photonic accelerator design incorporating a systematic analysis of practical scaling limits and introduces the Symmetric Grid Rule for improved utilization.

Findings

Topology-dominated scaling bottleneck identified as the Utilization Wall.

Symmetric topologies improve utilization by up to 6X.

Topology-aware scaling reduces memory access by over 40%.

Abstract

The rapid growth of deep neural networks (DNNs) has exposed fundamental limitations in electronic accelerators, where data movement dominates energy consumption, commonly referred to as the memory wall. Photonic accelerators offer a compelling alternative due to their inherent parallelism and high-speed matrix operations. However, existing research largely focuses on device-level innovations, leaving system-level scalability insufficiently explored. In this paper, we present a scalable photonic accelerator architecture based on a modular scale-out paradigm using 4 X 4 photonic tensor core units. We perform a systematic architectural analysis that incorporates the practical scaling limits of photonic hardware, including insertion loss, fanout penalties, and laser power limits, which restrict monolithic photonic scaling. Through evaluation on representative DNN workloads (GoogleNet, ResNet-18, MobileNet, and AlphaGo Zero) with up to 1024 processing elements, we identify a topology-dominated scaling bottleneck in the photonic domain, termed the Utilization Wall, where performance is governed by grid topology rather than hardware size. We further establish the Symmetric Grid Rule, demonstrating that symmetric topologies improve utilization by up to 6X while reducing memory access by over 40% compared to linear configurations, which reveal that topology-aware scaling is essential for achieving energy-efficient and high-performance photonic AI accelerators.