Processing

Please wait...

Settings

Settings

Goto Application

1. WO2020193728 - A PARTITIONABLE NETWORKED COMPUTER

Publication Number WO/2020/193728
Publication Date 01.10.2020
International Application No. PCT/EP2020/058618
International Filing Date 26.03.2020
IPC
G06F 15/173 2006.01
GPHYSICS
06COMPUTING; CALCULATING OR COUNTING
FELECTRIC DIGITAL DATA PROCESSING
15Digital computers in general; Data processing equipment in general
16Combinations of two or more digital computers each having at least an arithmetic unit, a program unit and a register, e.g. for a simultaneous processing of several programs
163Interprocessor communication
173using an interconnection network, e.g. matrix, shuffle, pyramid, star or snowflake
CPC
G06F 15/17306
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
FELECTRIC DIGITAL DATA PROCESSING
15Digital computers in general
16Combinations of two or more digital computers each having at least an arithmetic unit, a program unit and a register, e.g. for a simultaneous processing of several programs
163Interprocessor communication
173using an interconnection network, e.g. matrix, shuffle, pyramid, star, snowflake
17306Intercommunication techniques
G06F 15/17318
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
FELECTRIC DIGITAL DATA PROCESSING
15Digital computers in general
16Combinations of two or more digital computers each having at least an arithmetic unit, a program unit and a register, e.g. for a simultaneous processing of several programs
163Interprocessor communication
173using an interconnection network, e.g. matrix, shuffle, pyramid, star, snowflake
17306Intercommunication techniques
17318Parallel communications techniques, e.g. gather, scatter, reduce, roadcast, multicast, all to all
G06F 15/17375
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
FELECTRIC DIGITAL DATA PROCESSING
15Digital computers in general
16Combinations of two or more digital computers each having at least an arithmetic unit, a program unit and a register, e.g. for a simultaneous processing of several programs
163Interprocessor communication
173using an interconnection network, e.g. matrix, shuffle, pyramid, star, snowflake
17356Indirect interconnection networks
17368non hierarchical topologies
17375One dimensional, e.g. linear array, ring
G06F 15/17387
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
FELECTRIC DIGITAL DATA PROCESSING
15Digital computers in general
16Combinations of two or more digital computers each having at least an arithmetic unit, a program unit and a register, e.g. for a simultaneous processing of several programs
163Interprocessor communication
173using an interconnection network, e.g. matrix, shuffle, pyramid, star, snowflake
17356Indirect interconnection networks
17368non hierarchical topologies
17387Three dimensional, e.g. hypercubes
G06K 9/6256
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
9Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
62Methods or arrangements for recognition using electronic means
6217Design or setup of recognition systems and techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
6256Obtaining sets of training patterns; Bootstrap methods, e.g. bagging, boosting
G06N 20/00
GPHYSICS
06COMPUTING; CALCULATING; COUNTING
NCOMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
20Machine learning
Applicants
  • GRAPHCORE LIMITED [GB]/[GB]
Inventors
  • KNOWLES, Simon
Agents
  • PAGE WHITE & FARRER
Priority Data
1904265.427.03.2019GB
Publication Language English (EN)
Filing Language English (EN)
Designated States
Title
(EN) A PARTITIONABLE NETWORKED COMPUTER
(FR) ORDINATEUR EN RÉSEAU PARTAGEABLE
Abstract
(EN)
A computer is provided, comprising a plurality of processing nodes arranged two-dimensional arrays in respective front and rear layers. Each processing node has a set of activatable links. When activated, transmission of data items between the nodes connected via the activated link is enabled. When not activated, transmission of data items between said nodes is prevented. The set of activatable links comprises a respective link which connects the processing node to each adjacent node in the array, and to a facing processing node in the other layer. An allocation engine is configured to receive an allocation instruction and connected to the processing nodes to selectively activate the links in a configuration in which: (i) links between adjacent nodes are activated; (ii) links between facing nodes are only activated for edge processing nodes; and (iii) links between processing nodes outside the group and adjacent processing nodes in the group are deactivated.
(FR)
La présente invention concerne un ordinateur, comprenant une pluralité de nœuds de traitement agencés en réseaux en deux dimensions dans des couches avant et arrière respectives. Chaque nœud de traitement a un ensemble de liaisons activables. Lorsque ces dernières sont activées, la transmission d'éléments de données entre les nœuds connectés par l'intermédiaire de la liaison activée est activée. Lorsqu'elles ne sont pas activées, la transmission d'éléments de données entre lesdits nœuds est empêchée. L'ensemble de liaisons activables comprend une liaison respective qui connecte le nœud de traitement à chaque nœud adjacent dans le réseau, et à un nœud de traitement faisant face dans l'autre couche. Un moteur d'attribution est conçu pour recevoir une instruction d'attribution et est connecté aux nœuds de traitement pour activer sélectivement les liaisons dans une configuration dans laquelle : (i) des liaisons entre des nœuds adjacents sont activées ; (ii) des liaisons entre des nœuds se faisant face ne sont activées que pour des nœuds de traitement de bord ; et (iii) des liaisons entre des nœuds de traitement à l'extérieur du groupe et des nœuds de traitement adjacents dans le groupe sont désactivées.
Latest bibliographic data on file with the International Bureau