Abstract
We present an optical neural-network accelerator based on coherent (homodyne) detection that is scalable to large networks, and analyze the fundamental (quantum) and practical limits to its performance.
© 2019 The Author(s)
PDF ArticleMore Like This
Ryan Hamerly, Alex Sludds, Liane Bernstein, Marin Soljačić, and Dirk Englund
JF2F.5 CLEO: Applications and Technology (CLEO:A&T) 2019
Ruiting Wang, Pengfei Wang, Guangzhen Luo, Hongyan Yu, Xuliang Zhou, Yejin Zhang, and Jiaoqing Pan
T2D.4 Asia Communications and Photonics Conference (ACP) 2020
Xudong Zhao, Hui Yang, Huifeng Guo, Tao Peng, and Jie Zhang
M3J.5 Optical Fiber Communication Conference (OFC) 2019