-
公开(公告)号：US11209856B2
公开(公告)日：2021-12-28
申请号：US16799153
申请日：2020-02-24
Applicant: Lightmatter, Inc.
Inventor： Darius Bunandar , Martin B. Z. Forsythe , Michael Gould , Tomo Lazovich
IPC: G06E1/04
Abstract: Systems and methods for performing matrix operations using a path-number balanced optical network are provided. The optical network is formed as an array including active optical components and passive optical components arranged at a substantially central location of the array. The optical network includes at least NM active optical components which are used to implement a first matrix of any size N×M by embedding the first matrix in a second matrix of a larger size. The optical network performs matrix-vector and matrix-matrix operations by propagating one or more pluralities of optical signals corresponding to an input vector through the optical network.
-
公开(公告)号：US11023691B2
公开(公告)日：2021-06-01
申请号：US16995674
申请日：2020-08-17
Applicant: Lightmatter, Inc.
Inventor： Tyler J. Kenney , Martin B. Z. Forsythe , Tomo Lazovich , Darius Bunandar
Abstract: Techniques for computing matrix operations for arbitrarily large matrices on a finite-sized hybrid analog-digital matrix processor are described. Techniques for gain adjustment in a finite-sized hybrid analog-digital matrix processor are described which enable the system to obtain higher energy efficiencies, greater physical density and improved numerical accuracy. In some embodiments, these techniques enable maximization of the predictive accuracy of a GEMM-based convolutional neural network using low-precision data representations.
-
公开(公告)号：US20210125066A1
公开(公告)日：2021-04-29
申请号：US17081841
申请日：2020-10-27
Applicant: Lightmatter, Inc.
Inventor： Tomo Lazovich
Abstract: Described herein are techniques for determining an architecture of a machine learning model that optimizes the machine learning model. The system obtains a machine learning model configured with a first architecture of a plurality of architectures. The machine learning model has a first set of parameters. The system determines a second architecture using a quantization of the parameters of the machine learning model. The system updates the machine learning model to obtain a machine learning model configured with the second architecture.
-
公开(公告)号：US20200272795A1
公开(公告)日：2020-08-27
申请号：US16801015
申请日：2020-02-25
Applicant: Lightmatter, Inc.
Inventor： Tyler J. Kenney , Martin B.Z. Forsythe , Tomo Lazovich , Darius Bunandar
Abstract: Techniques for computing matrix operations for arbitrarily large matrices on a finite-sized hybrid analog-digital matrix processor are described. Techniques for gain adjustment in a finite-sized hybrid analog-digital matrix processor are described which enable the system to obtain higher energy efficiencies, greater physical density and improved numerical accuracy. In some embodiments, these techniques enable maximization of the predictive accuracy of a GEMM-based convolutional neural network using low-precision data representations.
-
公开(公告)号：US11709520B2
公开(公告)日：2023-07-25
申请号：US17507325
申请日：2021-10-21
Applicant: Lightmatter, Inc.
Inventor： Darius Bunandar , Martin B. Z. Forsythe , Michael Gould , Tomo Lazovich
IPC: G06E1/04
CPC classification number: G06E1/045
Abstract: Systems and methods for performing matrix operations using a path-number balanced optical network are provided. The optical network is formed as an array including active optical components and passive optical components arranged at a substantially central location of the array. The optical network includes at least NM active optical components which are used to implement a first matrix of any size N×M by embedding the first matrix in a second matrix of a larger size. The optical network performs matrix-vector and matrix-matrix operations by propagating one or more pluralities of optical signals corresponding to an input vector through the optical network.
-
Patent Agency Ranking
公开(公告)号：US20220416908A1
公开(公告)日：2022-12-29
申请号：US17840515
申请日：2022-06-14
Applicant: Lightmatter, Inc.
Inventor： Darius Bunandar , Nicholas C. Harris , Michael Gould , Carl Ramey , Tomo Lazovich
Abstract: Systems and methods for performing signed matrix operations using a linear photonic processor are provided. The linear photonic processor is formed as an array of first amplitude modulators and second amplitude modulators, the first amplitude modulators configured to encode elements of a vector into first optical signals and the second amplitude modulators configured to encode a product between the vector elements and matrix elements into second optical signals. An apparatus may be used to implement a signed value of an output of the linear processor. The linear photonic processor may be configured to perform matrix-vector and/or matrix-matrix operations.
-
公开(公告)号：US20220366308A1
公开(公告)日：2022-11-17
申请号：US17864172
申请日：2022-07-13
Applicant: Lightmatter, Inc.
Inventor： Tomo Lazovich , Darius Bunandar , Nicholas C. Harris , Martin B.Z. Forsythe
Abstract: Methods and apparatus for training a matrix-based differentiable program using a photonics-based processor. The matrix-based differentiable program includes at least one matrix-valued variable associated with a matrix of values in a Euclidean vector space. The method comprises configuring components of the photonics-based processor to represent the matrix of values as an angular representation, processing, using the components of the photonics-based processor, training data to compute an error vector, determining in parallel, at least some gradients of parameters of the angular representation, wherein the determining is based on the error vector and a current input training vector, and updating the matrix of values by updating the angular representation based on the determined gradients.
-
公开(公告)号：US20210279432A1
公开(公告)日：2021-09-09
申请号：US17246892
申请日：2021-05-03
Applicant: Lightmatter, Inc.
Inventor： TYLER J. KENNEY , Martin B.Z. Forsythe , Tomo Lazovich , Darius Bunandar
Abstract: Techniques for computing matrix operations for arbitrarily large matrices on a finite-sized hybrid analog-digital matrix processor are described. Techniques for gain adjustment in a finite-sized hybrid analog-digital matrix processor are described which enable the system to obtain higher energy efficiencies, greater physical density and improved numerical accuracy. In some embodiments, these techniques enable maximization of the predictive accuracy of a GEMM-based convolutional neural network using low-precision data representations.
-
公开(公告)号：US20210036783A1
公开(公告)日：2021-02-04
申请号：US16940900
申请日：2020-07-28
Applicant: Lightmatter, Inc.
Inventor： Darius Bunandar , Nicholas C. Harris , Michael Gould , Carl Ramey , Tomo Lazovich
Abstract: Systems and methods for performing signed matrix operations using a linear photonic processor are provided. The linear photonic processor is formed as an array of first amplitude modulators and second amplitude modulators, the first amplitude modulators configured to encode elements of a vector into first optical signals and the second amplitude modulators configured to encode a product between the vector elements and matrix elements into second optical signals. An apparatus may be used to implement a signed value of an output of the linear processor. The linear photonic processor may be configured to perform matrix-vector and/or matrix-matrix operations.
-
公开(公告)号：US20200380217A1
公开(公告)日：2020-12-03
申请号：US16995674
申请日：2020-08-17
Applicant: Lightmatter, Inc.
Inventor： Tyler J. Kenney , Martin B.Z. Forsythe , Tomo Lazovich , Darius Bunandar
Abstract: Techniques for computing matrix operations for arbitrarily large matrices on a finite-sized hybrid analog-digital matrix processor are described. Techniques for gain adjustment in a finite-sized hybrid analog-digital matrix processor are described which enable the system to obtain higher energy efficiencies, greater physical density and improved numerical accuracy. In some embodiments, these techniques enable maximization of the predictive accuracy of a GEMM-based convolutional neural network using low-precision data representations.
-
-
-
-
-
-
-
-
-
Search Results
33
records
(took 0.014 seconds)
