Photonic neural network convolutional layer chip based on micro-ring resonator

A technology of microring resonator and neural network, which is applied in the field of intelligent photon signal processing technology and neural network, can solve the problems of complex chips, affecting the accuracy of calculation, and the large number of cascade series of photon shifters, etc., and achieve high energy consumption ratio , Improving the feasibility of implementation and simplifying the effect of design complexity

Active Publication Date: 2019-04-16
SHANGHAI JIAO TONG UNIV
View PDF3 Cites 46 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the disadvantage of this method is also obvious. When the required calculation is more complicated, the chip realized by this metho

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Photonic neural network convolutional layer chip based on micro-ring resonator
  • Photonic neural network convolutional layer chip based on micro-ring resonator
  • Photonic neural network convolutional layer chip based on micro-ring resonator

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0037] The laser light source array 1 adopts InGaAsP / InP lasers in the embodiment.

[0038] The modulator array 2 in the embodiment adopts a silicon-based integrated electro-optic modulator.

[0039] Said wavelength division multiplexer 3 adopts arrayed waveguide grating in the embodiment.

[0040] Described microring resonator array adopts microring resonator cascaded structure 5.1 and double-balanced detection structure 5.4 of straight-through end 5.12 and coupling end 5.3 in the embodiment (see attached figure 2 ).

[0041] The laser light source array 1 has M (in this embodiment, M=4) optical output ports in total, and the modulator array 2 is composed of M optical input ports and M electrical input ports, M modulators and Consisting of M optical output ports, the wavelength division multiplexer 3 has M optical input ports and 1 optical output port, and the described optical splitter 4 has 1 optical input port and N (in this embodiment, N=3) optical output ports, the m...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a photonic neural network convolutional layer chip based on a micro-ring resonator. The chip is universal to all deep learning technologies including convolutional calculation.According to the chip, vectorized to-be-calculated signals are loaded on different optical wavelengths by means of a wavelength division multiplexing mode. The micro-ring resonator and a balance photoelectric detector form a weight matrix, convolutional calculation of the to-be-calculated signals and the weight matrix can be completed, and a convolutional result is output. By utilizing the tunability of the integrated micro-ring resonator, the convolutional calculation of any numerical value can be realized. In addition, the speed of the convolutional calculation is increased to a constant level (that is, the speed of light) by using light as a numerical calculation medium, and the advantage of a higher energy efficiency ratio is achieved.

Description

technical field [0001] The invention relates to intelligent photon signal processing technology and neural network technology, in particular to a photon neural network convolution layer chip. technical background [0002] In recent years, deep learning technology (Y. LeCun, et al, "Deep learning," Nature, vol.521, pp.436-444, 2015) has attracted widespread attention from academia and industry. Based on large-scale databases and high-speed digital computing capabilities, deep learning can automatically extract the key features of data, so as to realize tasks such as data prediction, reconstruction, and judgment. Several deep learning-based techniques have surpassed human performance in certain tasks and have been widely used as a result. However, an important condition for realizing high-performance deep learning is a digital computing platform with ultra-high speed and ultra-high energy efficiency. Existing digital computing platforms such as central processing units (CPUs...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): H04B10/50H04B10/54H04B10/572H04J14/02G06N3/067
CPCG06N3/0675H04B10/5053H04B10/54H04B10/572H04J14/0261
Inventor 邹卫文徐绍夫陈建平
Owner SHANGHAI JIAO TONG UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap