Paper
21 December 2023 An implementation of complex-valued matrix-vector multiplication based on MZI array for photonic neural networks
Zichao Zhao, Bei Chen, Ziyi Fu, Zeye Yu, Maohui Li, Yuehai Wang, Jianyi Yang
Author Affiliations +
Proceedings Volume 12966, AOPC 2023: AI in Optics and Photonics ; 1296621 (2023) https://doi.org/10.1117/12.3007862
Event: Applied Optics and Photonics China 2023 (AOPC2023), 2023, Beijing, China
Abstract
We propose a complex-valued matrix-vector multiplication method in this work, which make full use of the amplitude and phase information of input signal light and weight matrix. We demonstrate our computing method theoretically and experimentally. In order to verify our theory, a photonic chip is designed and used to setup a photonic neural network system, and the corresponding accuracies reach more than 90%.
(2023) Published by SPIE. Downloading of the abstract is permitted for personal use only.
Zichao Zhao, Bei Chen, Ziyi Fu, Zeye Yu, Maohui Li, Yuehai Wang, and Jianyi Yang "An implementation of complex-valued matrix-vector multiplication based on MZI array for photonic neural networks", Proc. SPIE 12966, AOPC 2023: AI in Optics and Photonics , 1296621 (21 December 2023); https://doi.org/10.1117/12.3007862
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Neural networks

Matrices

Matrix multiplication

Silicon

Continuous wave operation

Data acquisition

Education and training

Back to Top