A Test Methodology for Neural Computing Unit

Minho Cheong; Ingeol Lee; Sungho Kang

doi:10.1109/ISOCC.2018.8649896

A Test Methodology for Neural Computing Unit

Minho Cheong, Ingeol Lee, Sungho Kang

Department of Electrical and Electronic Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

2 Citations (Scopus)

Abstract

As convolutional neural networks (CNN) has been widely employed in deep learning applications, the accelerator for CNN has been proposed. Neural computing unit (NCU), which is an accelerator for CNN, includes thousands of identical cores named multiplier and accumulate (MAC), so testing NCU with the conventional methods are inefficient. This paper proposes a novel method to test NCU by applying test patterns for a MAC to all MACs in NCU. The experimental results indicate that the new method reduces test time to 1.38% and test data volume to 0.03%.

Original language	English
Title of host publication	Proceedings - International SoC Design Conference 2018, ISOCC 2018
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	11-12
Number of pages	2
ISBN (Electronic)	9781538679609
DOIs	https://doi.org/10.1109/ISOCC.2018.8649896
Publication status	Published - 2018 Jul 2
Event	15th International SoC Design Conference, ISOCC 2018 - Daegu, Korea, Republic of Duration: 2018 Nov 12 → 2018 Nov 15

Publication series

Name	Proceedings - International SoC Design Conference 2018, ISOCC 2018

Conference

Conference	15th International SoC Design Conference, ISOCC 2018
Country/Territory	Korea, Republic of
City	Daegu
Period	18/11/12 → 18/11/15

Bibliographical note

Publisher Copyright:
© 2018 IEEE.

All Science Journal Classification (ASJC) codes

Hardware and Architecture
Electrical and Electronic Engineering
Electronic, Optical and Magnetic Materials

Access to Document

10.1109/ISOCC.2018.8649896

Cite this

@inproceedings{f6948d1481d84137975742a8fc8cce83,

title = "A Test Methodology for Neural Computing Unit",

abstract = "As convolutional neural networks (CNN) has been widely employed in deep learning applications, the accelerator for CNN has been proposed. Neural computing unit (NCU), which is an accelerator for CNN, includes thousands of identical cores named multiplier and accumulate (MAC), so testing NCU with the conventional methods are inefficient. This paper proposes a novel method to test NCU by applying test patterns for a MAC to all MACs in NCU. The experimental results indicate that the new method reduces test time to 1.38% and test data volume to 0.03%.",

author = "Minho Cheong and Ingeol Lee and Sungho Kang",

note = "Publisher Copyright: {\textcopyright} 2018 IEEE.; 15th International SoC Design Conference, ISOCC 2018 ; Conference date: 12-11-2018 Through 15-11-2018",

year = "2018",

month = jul,

day = "2",

doi = "10.1109/ISOCC.2018.8649896",

language = "English",

series = "Proceedings - International SoC Design Conference 2018, ISOCC 2018",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "11--12",

booktitle = "Proceedings - International SoC Design Conference 2018, ISOCC 2018",

address = "United States",

}

Cheong, M, Lee, I & Kang, S 2018, A Test Methodology for Neural Computing Unit. in Proceedings - International SoC Design Conference 2018, ISOCC 2018., 8649896, Proceedings - International SoC Design Conference 2018, ISOCC 2018, Institute of Electrical and Electronics Engineers Inc., pp. 11-12, 15th International SoC Design Conference, ISOCC 2018, Daegu, Korea, Republic of, 18/11/12. https://doi.org/10.1109/ISOCC.2018.8649896

A Test Methodology for Neural Computing Unit. / Cheong, Minho; Lee, Ingeol; Kang, Sungho.
Proceedings - International SoC Design Conference 2018, ISOCC 2018. Institute of Electrical and Electronics Engineers Inc., 2018. p. 11-12 8649896 (Proceedings - International SoC Design Conference 2018, ISOCC 2018).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - A Test Methodology for Neural Computing Unit

AU - Cheong, Minho

AU - Lee, Ingeol

AU - Kang, Sungho

PY - 2018/7/2

Y1 - 2018/7/2

N2 - As convolutional neural networks (CNN) has been widely employed in deep learning applications, the accelerator for CNN has been proposed. Neural computing unit (NCU), which is an accelerator for CNN, includes thousands of identical cores named multiplier and accumulate (MAC), so testing NCU with the conventional methods are inefficient. This paper proposes a novel method to test NCU by applying test patterns for a MAC to all MACs in NCU. The experimental results indicate that the new method reduces test time to 1.38% and test data volume to 0.03%.

AB - As convolutional neural networks (CNN) has been widely employed in deep learning applications, the accelerator for CNN has been proposed. Neural computing unit (NCU), which is an accelerator for CNN, includes thousands of identical cores named multiplier and accumulate (MAC), so testing NCU with the conventional methods are inefficient. This paper proposes a novel method to test NCU by applying test patterns for a MAC to all MACs in NCU. The experimental results indicate that the new method reduces test time to 1.38% and test data volume to 0.03%.

UR - http://www.scopus.com/inward/record.url?scp=85063187348&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85063187348&partnerID=8YFLogxK

U2 - 10.1109/ISOCC.2018.8649896

DO - 10.1109/ISOCC.2018.8649896

M3 - Conference contribution

AN - SCOPUS:85063187348

T3 - Proceedings - International SoC Design Conference 2018, ISOCC 2018

SP - 11

EP - 12

BT - Proceedings - International SoC Design Conference 2018, ISOCC 2018

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 15th International SoC Design Conference, ISOCC 2018

Y2 - 12 November 2018 through 15 November 2018

ER -

A Test Methodology for Neural Computing Unit

Abstract

Publication series

Conference

Bibliographical note

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this