Constant depth, near constant depth, and subcubic size threshold circuits for linear algebraic calculations

Aimone, James Bradley; Parekh, Ojas D.; Phillips, Cynthia A.

Title: Constant depth, near constant depth, and subcubic size threshold circuits for linear algebraic calculations

Abstract

A method of increasing an efficiency at which a plurality of threshold gates arranged as neuromorphic hardware is able to perform a linear algebraic calculation having a dominant size of N. The computer-implemented method includes using the plurality of threshold gates to perform the linear algebraic calculation in a manner that is simultaneously efficient and at a near constant depth. “Efficient” is defined as a calculation algorithm that uses fewer of the plurality of threshold gates than a naïve algorithm. The naïve algorithm is a straightforward algorithm for solving the linear algebraic calculation. “Constant depth” is defined as an algorithm that has an execution time that is independent of a size of an input to the linear algebraic calculation. The near constant depth comprises a computing depth equal to or between O(log(log(N)) and the constant depth.

Inventors:: Aimone, James Bradley; Parekh, Ojas D.; Phillips, Cynthia A.

Issue Date:: 2019-10-15

Research Org.:: Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)

Sponsoring Org.:: USDOE

OSTI Identifier:: 1600161

Patent Number(s):: 10445065

Application Number:: 15/699,077

Assignee:: National Technology & Engineering Solutions of Sandia, LLC (Albuquerque, NM)

Patent Classifications (CPCs):: G - PHYSICS G06 - COMPUTING G06F - ELECTRIC DIGITAL DATA PROCESSING

G - PHYSICS G06 - COMPUTING G06N - COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS

Show more

G - PHYSICS G06 - COMPUTING G06F - ELECTRIC DIGITAL DATA PROCESSING
G06F17/12 - Simultaneous equations {, e.g. systems of linear equations}
G06F17/16 - Matrix or vector computation {, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
G06F2207/4818 - Threshold devices
G06F2207/4822 - Majority gates
G06F2207/4835 - Computations with rational numbers
G06F7/4833 - {Logarithmic number system}
G06F7/4876 - {Multiplying}
G06F7/5324 - {partitioned, i.e. using repetitively a smaller parallel parallel multiplier or using an array of such smaller multipliers}

G - PHYSICS G06 - COMPUTING G06N - COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
G06N3/049 - {Temporal neural nets, e.g. delay elements, oscillating neurons, pulsed inputs}
G06N3/063 - using electronic means

Show less

DOE Contract Number:: NA0003525

Resource Type:: Patent

Resource Relation:: Patent File Date: 09/08/2017

Country of Publication:: United States

Language:: English

Subject:: 97 MATHEMATICS AND COMPUTING

Citation Formats


                    Aimone, James Bradley, Parekh, Ojas D., and Phillips, Cynthia A.. Constant depth, near constant depth, and subcubic size threshold circuits for linear algebraic calculations.  United States: N. p., 2019. 
        Web.

Copy to clipboard


                    Aimone, James Bradley, Parekh, Ojas D., & Phillips, Cynthia A.. Constant depth, near constant depth, and subcubic size threshold circuits for linear algebraic calculations.  United States.

Copy to clipboard


                    Aimone, James Bradley, Parekh, Ojas D., and Phillips, Cynthia A.. Tue .  
        "Constant depth, near constant depth, and subcubic size threshold circuits for linear algebraic calculations".  United States.  https://www.osti.gov/servlets/purl/1600161.

Copy to clipboard


                    
@article{osti_1600161,

  title        = {Constant depth, near constant depth, and subcubic size threshold circuits for linear algebraic calculations},

  author       = {Aimone, James Bradley and Parekh, Ojas D. and Phillips, Cynthia A.},

  abstractNote = {A method of increasing an efficiency at which a plurality of threshold gates arranged as neuromorphic hardware is able to perform a linear algebraic calculation having a dominant size of N. The computer-implemented method includes using the plurality of threshold gates to perform the linear algebraic calculation in a manner that is simultaneously efficient and at a near constant depth. “Efficient” is defined as a calculation algorithm that uses fewer of the plurality of threshold gates than a naïve algorithm. The naïve algorithm is a straightforward algorithm for solving the linear algebraic calculation. “Constant depth” is defined as an algorithm that has an execution time that is independent of a size of an input to the linear algebraic calculation. The near constant depth comprises a computing depth equal to or between O(log(log(N)) and the constant depth.},

  doi          = {},

  journal      = {},
number       = ,

  volume       = ,

  place        = {United States},

  year         = {2019},

  month        = {10}

}

Copy to clipboard

Patent:

Save / Share:

Export Metadata

Save to My Library

Works referenced in this record:

Methods for Performing DAF Data Filtering and Padding
patent-application, April 2004

Kouri, Donald J.; Hoffman, David K.; Kakadiaris, Ioannis A.
US Patent Application 10/454373; 20040071363
URL: https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/20040071363

Method, Apparatus and Computer Program Providing Broadband Preconditioning Based on Reduced Coupling for Numerical Solvers
patent-application, October 2005

Feldmann, Peter; Morsey, Jason D.; Rubin, Barry J.
US Patent Application 10/815432; 20050222825
URL: https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/20050222825

Analysis Method and System
patent-application, October 2005

Cheng, Hongwei; Greengard, Leslie Frederick; Gimbutas, Zydrunas
US Patent Application 11/053159; 20050234686
URL: https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/20050234686

Similar Records in DOE Patents and OSTI.GOV collections:

Constant Depth and Subcubic Size Threshold Circuits for Matrix Multiplication.

Conference Parekh, Ojas D.; Phillips, Cynthia A.; Galiardi, Meghan; ...

Abstract not provided.
Full Text Available
Constant-Depth and Subcubic-Size Threshold Circuits for Matrix Multiplication.

Conference Parekh, Ojas D.; Phillips, Cynthia A.; James, Conrad D.; ...

Abstract not provided.
DOI: https://doi.org/10.1145/3210377.3210410

Full Text Available
Linear fitting of multi-threshold counting data

Patent Simpson, Garth Jason; Muir, Ryan Douglas; Pogranichniy, Nicholas Roman

The present disclosure provides a system and method for efficiently mining multi-threshold measurements acquired using photon counting pixel-array detectors for spectral imaging and diffraction analyses. Images of X-ray intensity as a function of X-ray energy were recorded on a 6 megapixel X-ray photon counting array detector through linear fitting of the measured counts recorded as a function of counting threshold. An analytical model is disclosed for describing the probability density of detected voltage, utilizing fractional photon counting to account for edge/corner effects from voltage plumes that spread across multiple pixels. Three-parameter fits to the model were independently performed for eachmore » « less
Full Text Available
Communications circuit including a linear quadratic estimator

Patent Ferguson, Dennis D.

A circuit includes a linear quadratic estimator (LQE) configured to receive a plurality of measurements a signal. The LQE is configured to weight the measurements based on their respective uncertainties to produce weighted averages. The circuit further includes a controller coupled to the LQE and configured to selectively adjust at least one data link parameter associated with a communication channel in response to receiving the weighted averages.
Full Text Available
Proton radiography based on near-threshold Cerenkov radiation

Patent van Bibber, Karl A [Berkeley, CA]; Dietrich, Frank S [Walnut Creek, CA]

A Cerenkov imaging system for charged particle radiography that determines the energy loss of the charged particle beam passing through an object. This energy loss information provides additional detail on target densities when used with traditional radiographic techniques like photon or x-ray radiography. In this invention a probe beam of 800 MeV to 50 GeV/c charged particles is passed through an object to be imaged, an imaging magnetic spectrometer, to a silicon aerogel Cerenkov radiator where the charged particles emitted Cerenkov light proportional to their velocity. At the same beam focal plane, a particle scintillator produces a light output proportionalmore » « less
Full Text Available

Similar Records

Title: Constant depth, near constant depth, and subcubic size threshold circuits for linear algebraic calculations

Abstract

Citation Formats

Methods for Performing DAF Data Filtering and Padding patent-application, April 2004

Method, Apparatus and Computer Program Providing Broadband Preconditioning Based on Reduced Coupling for Numerical Solvers patent-application, October 2005

Analysis Method and System patent-application, October 2005

Methods for Performing DAF Data Filtering and Padding
patent-application, April 2004

Method, Apparatus and Computer Program Providing Broadband Preconditioning Based on Reduced Coupling for Numerical Solvers
patent-application, October 2005

Analysis Method and System
patent-application, October 2005