11 TeraFLOPs per second photonic convolutional accelerator for deep learning optical neural networks

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 TeraFLOPS (floating point 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.

21 pages, 9 figures, 39 references

References in corpus (4)

• All-optical signal processing platforms for CMOS compatible integrated nonlinear optics
• Integrated micro-comb sources for quantum optical applications
• Attojoule Optoelectronics for Low-Energy Information Processing and Communications: a Tutorial Review
• Optical combs with a crystalline whispering gallery mode resonator

Cited by in corpus (6)

• High bandwidth temporal RF photonic signal processing with Kerr micro-combs: integration, fractional differentiation and Hilbert transforms
• Photonic Microwave and RF Channelizers using Kerr Micro-combs
• Soliton crystal Kerr microcombs for high-speed, scalable optical neural networks at 10 GigaOPs/s
• Optical data transmission field trial @ 44Tb/s with a 49GHz Kerr soliton crystal microcomb
• Hardware realization of the multiply and accumulate operation on radio-frequency signals with magnetic tunnel junctions
• High Performance Optical Filters Using Three Waveguide Coupled Sagnac Loop Reflectors