Apparatus and methods for optical neural network

US10268232B2 · US · B2

Patent metadata
FieldValue
Publication numberUS-10268232-B2
Application numberUS-201715612043-A
CountryUS
Kind codeB2
Filing dateJun 2, 2017
Priority dateJun 2, 2016
Publication dateApr 23, 2019
Grant dateApr 23, 2019

How to read this patent

A practical reading order for non-experts. Skip the full description unless you need deep technical detail.

  1. Title

    What the patent document calls the invention.

  2. Abstract

    A short plain-language summary of the technical disclosure.

  3. Assignees and inventors

    Who owns or filed the patent and who is credited as inventor.

  4. Key dates

    Filing, priority, publication, and grant dates set the timeline.

  5. First independent claim

    The legal scope of protection — read this for what is actually claimed.

  6. CPC / IPC classifications

    Technology tags used to group this patent with similar filings.

  7. Citations and related patents

    Prior art links and similar publications in this corpus.

Abstract

Official abstract text for this publication.

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.

First claim

Opening claim text (preview).

The invention claimed is: 1. An apparatus for implementing an artificial neural network, the apparatus comprising: an array of input waveguides to receive a first array of optical signals; an optical interference unit, in optical communication with the array of input waveguides, to perform a linear transformation of the first array of optical signals into a second array of optical signals; an array of output waveguides, in optical communication with the optical interference unit, to guide the second array of optical signals, wherein at least one input waveguide in the array of input waveguides is in optical communication with each output waveguide in the array of output waveguides via the optical interference unit, a nonlinearity unit, in communication with the optical interference unit, to perform a nonlinear transformation on the second array of optical signals so as to generate a third array of signals, a light source in optical communication with the array of input waveguides; and feedback circuitry to transmit an array of electrical signals to the light source based on the third array of signals, the light source emitting a fourth array of optical signals based on the array of electrical signals. 2. The apparatus of claim 1 , wherein the optical interference unit comprises: a plurality of interconnected Mach-Zehnder interferometers (MZIs), each MZI in the plurality of interconnected MZIs comprising: a first phase shifter configured to change a splitting ratio of the MZI; and a second phase shifter configured to shift a phase of one output of the MZI. 3. The apparatus of claim 2 , wherein the plurality of interconnected MZIs is configured to perform the linear transformation of the first array of optical signals via singular value decomposition (SVD). 4. The apparatus of claim 2 , wherein the plurality of interconnected MZIs comprises: a first set of MZIs to perform a unitary transformation of the first array of optical signals to generate a first array of transformed optical signals; and a second set of MZIs in optical communication with the first set of MZIs, each MZI in the second array of MZIs receiving a respective transformed optical signal from the first array of transformed optical signals and transmitting a first output to the nonlinearity unit, wherein a second output of each MZI in the second array of MZIs is optically blocked so as to change a total intensity of the first array of transformed optical signals and perform a non-unitary diagonal matrix multiplication of the first array of transformed optical signals. 5. The apparatus of claim 1 , wherein the optical interference unit comprises: a plurality of interconnected MZIs to perform a unitary transformation of the first array of optical signals; and an array of optical attenuators or amplifiers, in optical communication with the plurality of MZIs, to change an intensity of the first array of optical signals after the plurality of MZIs. 6. The apparatus of claim 1 , wherein the optical interference unit comprises a network of photonic crystals. 7. The apparatus of claim 6 , further comprising a phase change material coated on at least one photonic crystal in the network of photonic crystals. 8. The apparatus of claim 1 , wherein the optical interference unit comprises a two-dimensional (2D) array of ring resonators. 9. The apparatus of claim 1 , wherein the first array of optical signals includes a first number of optical signals, the second array of the optical signals includes a second number of optical signals, and the first number is different than the second number. 10. The apparatus of claim 1 , wherein the third array of signals is a third array of optical signals, and wherein the nonlinearity unit is an optical nonlinearity unit, in optical communication with the optical interference unit, to perform the nonlinear transformation on the second array of optical signals so as to generate the third array of optical signals. 11. The apparatus of claim 10 , wherein the optical nonlinearity unit comprises an array of saturable absorbers, each saturable absorber in the array of saturable absorbers configured to absorb a corresponding optical signal in the second array of optical signal. 12. The apparatus of claim 10 , wherein the optical nonlinearity unit comprises an array of bi-stable optical switches, each bi-stable optical switch in the array of bi-stable optical receives a corresponding optical signal in the second array of optical signals. 13. The apparatus of claim 10 , wherein the optical nonlinearity unit comprises an array of ring resonators, each ring resonator in the array of ring resonators receiving a corresponding second optical signal from the interference unit. 14. The apparatus of claim 10 , further comprising: a detector array, in optical communication with the optical nonlinearity unit, to detect the third array of optical signals and generate the array of electrical signals. 15. The apparatus of claim 14 , further comprising: control circuitry, operably coupled to the optical interference unit, the optical nonlinearity unit, and the detector array, to adjust a setting of at least one of the optical interference unit or the optical nonlinearity unit based at least in part on the third array of optical signals detected by the detector array. 16. The apparatus of claim 1 , wherein the nonlinearity unit implements the nonlinear transformation in the electronic domain by measuring optical intensities of the second array of optical signals with a photodetector array and electronically applying the nonlinear transformation to the measured optical intensities to generate the third array of signals. 17. The apparatus of claim 1 , wherein the second array of optical signals can be represented as Z i =|E i | 2 , where E i is the electric field of i th optical signal in the second array of optical signals and the nonlinear transformation applies a nonlinear function f to the second array of optical signals to yield the third array of signals f(Z i ). 18. The apparatus of claim 1 , wherein the optical interference unit and the nonlinearity unit form one hidden layer in a sequence of hidden layers of the artificial neural network. 19. A method for artificial neural network computation, the method comprising: receiving a first array of optical signals with an array of input waveguides; interfering the first array of optical signals, using an optical interference unit in optical communication with the array of input waveguides, to linearly transform the first array of optical signals into a second array of optical signals; guiding the second array of optical signals using an array of output waveguides, wherein at least one input waveguide in the array of input waveguides is in optical communication with each output waveguide in the array of output waveguides via the optical interference unit, nonlinearly transforming the second array of optical signals using a nonlinearity unit, in communication with the optical interference unit, so as to generate a third array of signals; and providing a fourth array of optical signals to the array of input waveguides based on the third array of signals. 20. The method of claim 19 , wherein interfering the first array of optical signals comprises: propagating the first array of optical signals through a plurality of interconnected Mach-Zehnder interferometers (MZIs); changing a splitting ratio of at least one MZI in the plurality of interconnected MZIs; and shifting a phase of on

Assignees

Inventors

Classifications

  • G02F1/225Primary

    in an optical waveguide structure · CPC title

  • involving resonance effects, e.g. resonantly enhanced interaction · CPC title

  • Photonic crystals · CPC title

  • using two-photon emission or absorption processes · CPC title

  • Backpropagation, e.g. using gradient descent · CPC title

Patent family

Related publications grouped by family.

External sources

Frequently asked questions

Answers are generated from the same data shown on this page.

What does patent US10268232B2 cover?
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, whic…
Who is the assignee on this patent?
Massachusetts Inst Technology
What technology area does this patent fall under?
Primary CPC classification G02F1/225. Mapped technology areas include Physics.
When was this patent published?
Publication date Tue Apr 23 2019 00:00:00 GMT+0000 (Coordinated Universal Time) (B2). Legal status and post-grant events are not shown on this page.
What related patents are in patentsdb?
We list 10 related publications on this page (citations in our corpus or others sharing the same primary CPC).