Apparatus and methods for optical neural network

US11334107B2 · kind B2 · utility

Assignee

• Massachusetts Institute of Technology · US

Inventors

• Jacques Johannes Carolan · Cambridge, US
• Mihika Prabhu · Cambridge, US
• Scott A. Skirlo · Boston, US
• Yichen Shen · Hangzhou City, CN
• Marin Soljacic · Belmont, US
• Dirk Robert Englund · San Diego, US
• Nicholas C. Harris · Menlo Park, US

Key dates

• Filing date: Aug 6, 2020
• Grant date: May 17, 2022
• Expiry date: Nov 20, 2040

Classification

• Technology area (CPC G): Physics
• CPC primary: G02F2203/15
• WIPO field: Computer technology
• WIPO sector: Electrical engineering

Abstract

An optical neural network is constructed based on photonic integrated circuits to perform neuromorphic computing. In the optical neural network, matrix multiplication is implemented using one or more optical interference units, which can apply an arbitrary weighting matrix multiplication to an array of input optical signals. Nonlinear activation is realized by an optical nonlinearity unit, which can be based on nonlinear optical effects, such as saturable absorption. These calculations are implemented optically, thereby resulting in high calculation speeds and low power consumption in the optical neural network.