Optical neuromorphic processing at Tera-OP/s speeds based on Kerr soliton crystal microcombs

TL;DR

This paper demonstrates a high-speed optical neural network accelerator capable of performing over 10 Tera-Operations per second, enabling real-time image recognition with high efficiency and scalability for complex tasks.

Contribution

The authors present a universal optical vector convolutional accelerator using microcomb sources, achieving unprecedented processing speeds and demonstrating practical neural network applications.

Findings

Operates beyond 10 TOPS for image convolutions

Successfully recognizes 10 handwritten digits with 88% accuracy

Scalable approach suitable for complex neural network tasks

Abstract

Convolutional neural networks (CNNs), inspired by biological visual cortex systems, are a powerful category of artificial neural networks that can extract the hierarchical features of raw data to greatly reduce the network parametric complexity and enhance the predicting accuracy. They are of significant interest for machine learning tasks such as computer vision, speech recognition, playing board games and medical diagnosis. Optical neural networks offer the promise of dramatically accelerating computing speed to overcome the inherent bandwidth bottleneck of electronics. Here, we demonstrate a universal optical vector convolutional accelerator operating beyond 10 TeraOPS (TOPS: operations per second), generating convolutions of images of 250,000 pixels with 8 bit resolution for 10 kernels simultaneously, enough for facial image recognition. We then use the same hardware to sequentially form a deep optical CNN with ten output neurons, achieving successful recognition of full 10 digits with 900 pixel handwritten digit images with 88% accuracy. Our results are based on simultaneously interleaving temporal, wavelength and spatial dimensions enabled by an integrated microcomb source. This approach is scalable and trainable to much more complex networks for demanding applications such as unmanned vehicle and real time video recognition.