Processing

Please wait...

Settings

Settings

Goto Application

1. WO2021038182 - REFACTORING MAC COMPUTATIONS FOR REDUCED PROGRAMMING STEPS

Publication Number WO/2021/038182
Publication Date 04.03.2021
International Application No. PCT/GB2020/050720
International Filing Date 19.03.2020
IPC
G06F 7/58 2006.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
FELECTRIC DIGITAL DATA PROCESSING
7Methods or arrangements for processing data by operating upon the order or content of the data handled
58Random or pseudo-random number generators
G06F 7/544 2006.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
FELECTRIC DIGITAL DATA PROCESSING
7Methods or arrangements for processing data by operating upon the order or content of the data handled
38Methods or arrangements for performing computations using exclusively denominational number representation, e.g. using binary, ternary, decimal representation
48using non-contact-making devices, e.g. tube, solid state device; using unspecified devices
544for evaluating functions by calculation
CPC
G06F 17/16
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
FELECTRIC DIGITAL DATA PROCESSING
17Digital computing or data processing equipment or methods, specially adapted for specific functions
10Complex mathematical operations
16Matrix or vector computation ; , e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
G06F 2207/4814
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
FELECTRIC DIGITAL DATA PROCESSING
2207Indexing scheme relating to methods or arrangements for processing data by operating upon the order or content of the data handled
38Indexing scheme relating to groups G06F7/38 - G06F7/575
48Indexing scheme relating to groups G06F7/48 - G06F7/575
4802Special implementations
4814Non-logic devices, e.g. operational amplifiers
G06F 7/4876
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
FELECTRIC DIGITAL DATA PROCESSING
7Methods or arrangements for processing data by operating upon the order or content of the data handled
38Methods or arrangements for performing computations using exclusively denominational number representation, e.g. using binary, ternary, decimal representation
48using non-contact-making devices, e.g. tube, solid state device; using unspecified devices
483Computations with numbers represented by a non-linear combination of denominational numbers, e.g. rational numbers, logarithmic number system, floating-point numbers
487Multiplying; Dividing
4876Multiplying
G06F 7/5443
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
FELECTRIC DIGITAL DATA PROCESSING
7Methods or arrangements for processing data by operating upon the order or content of the data handled
38Methods or arrangements for performing computations using exclusively denominational number representation, e.g. using binary, ternary, decimal representation
48using non-contact-making devices, e.g. tube, solid state device; using unspecified devices
544for evaluating functions by calculation
5443Sum of products
G06F 9/30014
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
FELECTRIC DIGITAL DATA PROCESSING
9Arrangements for program control, e.g. control units
06using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
30Arrangements for executing machine instructions, e.g. instruction decode
30003Arrangements for executing specific machine instructions
30007to perform operations on data operands
3001Arithmetic instructions
30014with variable precision
G06F 9/3893
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
FELECTRIC DIGITAL DATA PROCESSING
9Arrangements for program control, e.g. control units
06using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
30Arrangements for executing machine instructions, e.g. instruction decode
38Concurrent instruction execution, e.g. pipeline, look ahead
3885using a plurality of independent parallel functional units
3893controlled in tandem, e.g. multiplier-accumulator
Applicants
  • ARM LIMITED [GB]/[GB]
Inventors
  • MATTINA, Matthew
  • DAS, Shidhartha
  • ROSENDALE, Glen Arnold
  • GARCIA REDONDO, Fernando
Agents
  • TLIP LTD
Priority Data
16/556,10129.08.2019US
Publication Language English (EN)
Filing Language English (EN)
Designated States
Title
(EN) REFACTORING MAC COMPUTATIONS FOR REDUCED PROGRAMMING STEPS
(FR) REFACTORISATION DE CALCULS MAC POUR DES ÉTAPES DE PROGRAMMATION RÉDUITES
Abstract
(EN)
A method and architecture for performing multiply-accumulate operations in a neural network is disclosed. The architecture includes a crossbar having a plurality of non-volatile memory elements. A plurality of input activations is applied to the crossbar, which are then summed by binary weight encoding a plurality of the non-volatile memory elements to connect the input activations to weight values. At least one of the plurality of non-volatile memory elements is then precision programmed.
(FR)
L'invention concerne un procédé et une architecture pour effectuer des opérations de multiplication-accumulation dans un réseau de neuronal. L'architecture comprend une barre transversale ayant une pluralité d'éléments de mémoire non volatile. Une pluralité d'activations d'entrée est appliquée à la barre transversale, celles-ci étant ensuite totalisées par un poids binaire codant une pluralité d'éléments de mémoire non volatile pour connecter les activations d'entrée à des valeurs de poids. Au moins l'un de la pluralité d'éléments de mémoire non volatile est ensuite programmé de manière précise.
Also published as
Latest bibliographic data on file with the International Bureau