Featured Discovery

Home > Featured Discovery > Prof. Zou Weiwen’s Team Made Breakthrough in Optical Intelligent Computing

Prof. Zou Weiwen’s Team Made Breakthrough in Optical Intelligent Computing

November 10, 2021      Author:

Recently, Prof. Zou Weiwen's team from the State Key Laboratory of Advanced Optical Communication Systems and Networks, SJTU School of Electronic Information and Electrical Engineering made a breakthrough in optical intelligent computing, with research results published in Light: Science & Applications, an influential journal in optics, under the title "Optical coherent dot-product chip for sophisticated deep learning regression". The team developed an optical coherent dot-product chip with the ability to operate complex neural networks, becoming the world's first to realize high-precision medical image reconstruction with an optical intelligent computing chip.

Using medical image reconstruction as a validation, Prof. Zou Weiwen's team run the AUTOMAP (for general-purpose image reconstruction) neural network model on the chip successfully, and the reconstruction quality reached an ideal level that was close to the performance of a 32-bit computer. This work not only helps address the difficulties in application of optical neural network, but also brings new ideas to the next generation of intelligent computing technology. Further improving the chip's device integration scale is likely to produce optical neural network processors with higher speed and lower power consumption, alleviating the contradiction between the upsurging demand for intelligent computing power and the limited computing power of traditional hardware.

Paper Link:https://www.nature.com/articles/s41377-021-00666-8

 

Abstract

Optical implementations of neural networks (ONNs) herald the next-generation high-speed and energy-efficient deep learning computing by harnessing the technical advantages of large bandwidth and high parallelism of optics. However, due to the problems of the incomplete numerical domain, limited hardware scale, or inadequate numerical accuracy, the majority of existing ONNs were studied for basic classification tasks. Given that regression is a fundamental form of deep learning and accounts for a large part of current artificial intelligence applications, it is necessary to master deep learning regression for further development and deployment of ONNs. Here, we demonstrate a silicon-based optical coherent dot-product chip (OCDC) capable of completing deep learning regression tasks. The OCDC adopts optical fields to carry out operations in the complete real-value domain instead of in only the positive domain. Via reusing, a single chip conducts matrix multiplications and convolutions in neural networks of any complexity. Also, hardware deviations are compensated via in-situ backpropagation control provided the simplicity of chip architecture. Therefore, the OCDC meets the requirements for sophisticated regression tasks and we successfully demonstrate a representative neural network, the AUTOMAP (a cutting-edge neural network model for image reconstruction). The quality of reconstructed images by the OCDC and a 32-bit digital computer is comparable. To the best of our knowledge, there is no precedent of performing such state-of-the-art regression tasks on ONN chips. It is anticipated that the OCDC can promote the novel accomplishment of ONNs in modern AI applications including autonomous driving, natural language processing, and scientific study.

 

Source:School of Electronic Information and Electrical Engineering, SJTU

Translated by Chen Chen

Proofread by Xiao Yangning