Control-free and efficient integrated photonic neural networks via hardware-aware training and pruning

TL;DR

This paper introduces a hardware-aware training and pruning method for integrated photonic neural networks, significantly improving accuracy and energy efficiency without complex control mechanisms, thus advancing practical large-scale photonic AI systems.

Contribution

The paper presents a novel training and pruning approach that enhances noise robustness and energy efficiency in photonic neural networks without requiring complex control hardware.

Findings

Achieved a 4-bit improvement in PNN accuracy

Increased handwritten digit classification accuracy from 67% to 95%

Reduced energy consumption by tenfold

Abstract

Integrated photonic neural networks (PNNs) are at the forefront of AI computing, leveraging on light's unique properties, such as large bandwidth, low latency, and potentially low power consumption. Nevertheless, the integrated optical components within PNNs are inherently sensitive to external disturbances and thermal interference, which can detrimentally affect computing accuracy and reliability. Current solutions often use complicated control methods, resulting in high hardware complexity impractical for large-scale PNNs. In response, we propose a novel hardware-aware training and pruning approach. The core idea is to train the parameters of a physical neural network towards its noise-robust and energy-efficient region. This innovation enables control-free and energy-efficient photonic computing. Our method is validated across diverse integrated PNN architectures. Through experimental validation, our approach significantly enhances the computing precision of MRR-based PNN, achieving a notable 4-bit improvement without the need for complex device control mechanisms or energy-intensive temperature stabilization circuits. Specifically, it improves the accuracy of experimental handwritten digit classification from 67.0% to 95.0%, nearing theoretical limits and achieved without a thermoelectric controller. Additionally, this approach reduces the energy by tenfold. We further extend the validation to various architectures, such as PCM-based PNN, demonstrating the broad applicability of our approach across different platforms. This advancement represents a significant step towards the practical, energy-efficient, and noise-resilient implementation of large-scale integrated PNNs.