JP7361192B2 - ハードウェア・アクセラレータとインターフェースするための方法 - Google Patents

ハードウェア・アクセラレータとインターフェースするための方法 Download PDF

Info

Publication number
JP7361192B2
JP7361192B2 JP2022500757A JP2022500757A JP7361192B2 JP 7361192 B2 JP7361192 B2 JP 7361192B2 JP 2022500757 A JP2022500757 A JP 2022500757A JP 2022500757 A JP2022500757 A JP 2022500757A JP 7361192 B2 JP7361192 B2 JP 7361192B2
Authority
JP
Japan
Prior art keywords
operations
routine
hardware accelerator
computational task
subset
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
JP2022500757A
Other languages
English (en)
Japanese (ja)
Other versions
JP2022541144A5 (https=
JP2022541144A (ja
Inventor
ピヴェトー、クリストフ
イオアンノウ、ニコラス
クラウチェック、イゴール
ガロ－ブルドー、マニュエル レ
セバスティアン、アブ
エレフセリウー、エヴァンゲロス、スタブロス
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
International Business Machines Corp
Original Assignee
International Business Machines Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by International Business Machines Corp filed Critical International Business Machines Corp
Publication of JP2022541144A publication Critical patent/JP2022541144A/ja
Publication of JP2022541144A5 publication Critical patent/JP2022541144A5/ja
Application granted granted Critical
Publication of JP7361192B2 publication Critical patent/JP7361192B2/ja
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
    • G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
    • G06F9/30003—Arrangements for executing specific machine instructions
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/48—Program initiating; Program switching, e.g. by interrupt
    • G06F9/4806—Task transfer initiation or dispatching
    • G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
    • G06F9/4881—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
    • G06N3/063—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using electronic means
    • G06N3/065—Analogue means
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2209/00—Indexing scheme relating to G06F9/00
    • G06F2209/50—Indexing scheme relating to G06F9/50
    • G06F2209/509—Offload

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computing Systems (AREA)
  • Molecular Biology (AREA)
  • General Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Computational Linguistics (AREA)
  • Artificial Intelligence (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Neurology (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • Complex Calculations (AREA)
  • Advance Control (AREA)
JP2022500757A 2019-07-15 2020-06-30 ハードウェア・アクセラレータとインターフェースするための方法 Active JP7361192B2 (ja)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
US16/511,689 US11250107B2 (en) 2019-07-15 2019-07-15 Method for interfacing with hardware accelerators
US16/511,689 2019-07-15
PCT/EP2020/068377 WO2021008868A1 (en) 2019-07-15 2020-06-30 A method for interfacing with hardware accelerators

Publications (3)

Publication Number Publication Date
JP2022541144A JP2022541144A (ja) 2022-09-22
JP2022541144A5 JP2022541144A5 (https=) 2022-11-18
JP7361192B2 true JP7361192B2 (ja) 2023-10-13

Family

ID=71409414

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2022500757A Active JP7361192B2 (ja) 2019-07-15 2020-06-30 ハードウェア・アクセラレータとインターフェースするための方法

Country Status (5)

Country Link
US (1) US11250107B2 (https=)
EP (1) EP3999957B1 (https=)
JP (1) JP7361192B2 (https=)
CN (1) CN114127689B (https=)
WO (1) WO2021008868A1 (https=)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3933709A1 (en) * 2020-06-30 2022-01-05 Upstride Graph processing method and system
EP4285286A1 (en) * 2021-02-01 2023-12-06 Microsoft Technology Licensing, LLC Semi-programmable and reconfigurable co-accelerator for a deep neural network with normalization or non-linearity
CN115904696A (zh) * 2021-09-30 2023-04-04 想象技术有限公司 用于配置具有可配置流水线的神经网络加速器的方法和设备
TR2021020689A2 (tr) * 2021-12-22 2023-07-21 Havelsan Hava Elektronik Sanayi Ve Ticaret Anonim Sirketi Gömülü ve bütünleşi̇k si̇stemlerde paralel yapay si̇ni̇r ağlari i̇le topluluk öğrenmesi̇
US20240161222A1 (en) * 2022-11-16 2024-05-16 Nvidia Corporation Application programming interface to indicate image-to-column transformation
US12455900B1 (en) 2023-03-07 2025-10-28 QEngine LLC Method for executing a query in a multi-dimensional data space using vectorization and a related system
KR102740239B1 (ko) * 2023-03-24 2024-12-10 한국과학기술원 다중 신경망 가속을 위한 확장가능 벡터-어레이 이종 가속기 구조 및 스케쥴링 기법

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090217275A1 (en) 2008-02-22 2009-08-27 International Business Machines Corporation Pipelining hardware accelerators to computer systems
JP2012256150A (ja) 2011-06-08 2012-12-27 Renesas Electronics Corp コンパイル装置、コンパイル方法、及びプログラム
US20150006341A1 (en) 2013-06-27 2015-01-01 Metratech Corp. Billing transaction scheduling
JP2017138867A (ja) 2016-02-05 2017-08-10 富士通株式会社 演算処理回路、および情報処理装置
JP2018173672A (ja) 2015-09-03 2018-11-08 株式会社Ｐｒｅｆｅｒｒｅｄ Ｎｅｔｗｏｒｋｓ 実装装置
JP2019003546A (ja) 2017-06-19 2019-01-10 株式会社デンソー 多層ニューラルネットワークのニューロンの出力レベル調整方法
US20190114529A1 (en) 2017-10-17 2019-04-18 Xilinx, Inc. Multi-layer neural network processing by a neural network accelerator using host communicated merged weights and a package of per-layer instructions

Family Cites Families (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6081890A (en) * 1998-11-30 2000-06-27 Intel Corporation Method of communication between firmware written for different instruction set architectures
US7219085B2 (en) * 2003-12-09 2007-05-15 Microsoft Corporation System and method for accelerating and optimizing the processing of machine learning techniques using a graphics processing unit
US7984267B2 (en) * 2008-09-04 2011-07-19 International Business Machines Corporation Message passing module in hybrid computing system starting and sending operation information to service program for accelerator to execute application program
US8751556B2 (en) 2010-06-11 2014-06-10 Massachusetts Institute Of Technology Processor for large graph algorithm computations and matrix operations
US8752064B2 (en) * 2010-12-14 2014-06-10 Advanced Micro Devices, Inc. Optimizing communication of system call requests
US9411853B1 (en) 2012-08-03 2016-08-09 Healthstudio, LLC In-memory aggregation system and method of multidimensional data processing for enhancing speed and scalability
US9471388B2 (en) * 2013-03-14 2016-10-18 Altera Corporation Mapping network applications to a hybrid programmable many-core device
US10540588B2 (en) 2015-06-29 2020-01-21 Microsoft Technology Licensing, Llc Deep neural network processing on hardware accelerators with stacked memory
US9646243B1 (en) 2016-09-12 2017-05-09 International Business Machines Corporation Convolutional neural networks using resistive processing unit array
US10698766B2 (en) * 2018-04-18 2020-06-30 EMC IP Holding Company LLC Optimization of checkpoint operations for deep learning computing
CN108876702A (zh) * 2018-06-21 2018-11-23 北京邮电大学 一种加速分布式深度神经网络的训练方法及装置
US10620951B2 (en) * 2018-06-22 2020-04-14 Intel Corporation Matrix multiplication acceleration of sparse matrices using column folding and squeezing

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090217275A1 (en) 2008-02-22 2009-08-27 International Business Machines Corporation Pipelining hardware accelerators to computer systems
JP2012256150A (ja) 2011-06-08 2012-12-27 Renesas Electronics Corp コンパイル装置、コンパイル方法、及びプログラム
US20150006341A1 (en) 2013-06-27 2015-01-01 Metratech Corp. Billing transaction scheduling
JP2018173672A (ja) 2015-09-03 2018-11-08 株式会社Ｐｒｅｆｅｒｒｅｄ Ｎｅｔｗｏｒｋｓ 実装装置
JP2017138867A (ja) 2016-02-05 2017-08-10 富士通株式会社 演算処理回路、および情報処理装置
US20170228634A1 (en) 2016-02-05 2017-08-10 Fujitsu Limited Arithmetic processing circuit and information processing apparatus
JP2019003546A (ja) 2017-06-19 2019-01-10 株式会社デンソー 多層ニューラルネットワークのニューロンの出力レベル調整方法
US20200110991A1 (en) 2017-06-19 2020-04-09 Denso Corporation Method for adjusting output level of multilayer neural network neuron
US20190114529A1 (en) 2017-10-17 2019-04-18 Xilinx, Inc. Multi-layer neural network processing by a neural network accelerator using host communicated merged weights and a package of per-layer instructions
JP2020537785A (ja) 2017-10-17 2020-12-24 ザイリンクス インコーポレイテッドＸｉｌｉｎｘ Ｉｎｃｏｒｐｏｒａｔｅｄ ホスト通信されるマージされた重みと層単位命令のパッケージとを使用するニューラルネットワークアクセラレータによる多層ニューラルネットワーク処理

Also Published As

Publication number Publication date
JP2022541144A (ja) 2022-09-22
WO2021008868A1 (en) 2021-01-21
US20210019362A1 (en) 2021-01-21
US11250107B2 (en) 2022-02-15
EP3999957A1 (en) 2022-05-25
CN114127689B (zh) 2025-06-06
CN114127689A (zh) 2022-03-01
EP3999957B1 (en) 2025-09-24

Similar Documents

Publication Publication Date Title
JP7361192B2 (ja) ハードウェア・アクセラレータとインターフェースするための方法
US10713568B2 (en) Apparatus and method for executing reversal training of artificial neural network
US10839292B2 (en) Accelerated neural network training using a pipelined resistive processing unit architecture
CN107862374B (zh) 基于流水线的神经网络处理系统和处理方法
US11880768B2 (en) Method and apparatus with bit-serial data processing of a neural network
US10553207B2 (en) Systems and methods for employing predication in computational models
KR20180070103A (ko) 인식 방법 및 인식 장치
US11562235B2 (en) Activation function computation for neural networks
CN113454648B (zh) 循环神经网络中的勒让德存储器单元
CN111971693B (zh) 神经推理处理器的中央调度器和指令分配器
US20230297819A1 (en) Processor array for processing sparse binary neural networks
KR20190055408A (ko) 시냅스소자에서 가중치에 대한 비대칭성을 최소화하는 펄스 구동 장치 및 그 방법
KR102396447B1 (ko) 파이프라인 구조를 가지는 인공신경망용 연산 가속 장치
JP2022541144A5 (https=)
EP3803580B1 (en) Efficient incident management in large scale computer systems
Domingos et al. An efficient and scalable architecture for neural networks with backpropagation learning
CN114556373A (zh) 用于大规模并行神经推理引擎的多模式低精度内积计算电路
US11586895B1 (en) Recursive neural network using random access memory
CN110874633A (zh) 具有多位神经形态操作的神经形态方法和设备
KR20230029759A (ko) 아날로그 크로스바 어레이들을 업데이트하기 위한 희소 수정가능 비트 길이 결정 펄스 생성
KR102090109B1 (ko) 학습 및 추론 장치 및 그 방법
US20210142153A1 (en) Resistive processing unit scalable execution
US20230196124A1 (en) Runtime predictors for neural network computation reduction
CN112633464A (zh) 用于识别图像的计算系统和方法
US20250355711A1 (en) Re-rounding in integrated circuit for variance reduction in ai operations

Legal Events

Date Code Title Description
RD04 Notification of resignation of power of attorney

Free format text: JAPANESE INTERMEDIATE CODE: A7424

Effective date: 20220518

A521 Request for written amendment filed

Free format text: JAPANESE INTERMEDIATE CODE: A523

Effective date: 20221109

A621 Written request for application examination

Free format text: JAPANESE INTERMEDIATE CODE: A621

Effective date: 20221121

A977 Report on retrieval

Free format text: JAPANESE INTERMEDIATE CODE: A971007

Effective date: 20230823

TRDD Decision of grant or rejection written
A01 Written decision to grant a patent or to grant a registration (utility model)

Free format text: JAPANESE INTERMEDIATE CODE: A01

Effective date: 20230912

A61 First payment of annual fees (during grant procedure)

Free format text: JAPANESE INTERMEDIATE CODE: A61

Effective date: 20231002

R150 Certificate of patent or registration of utility model

Ref document number: 7361192

Country of ref document: JP

Free format text: JAPANESE INTERMEDIATE CODE: R150