IEICE Technical Committee Submission System
Conference Paper's Information
Online Proceedings
[Sign in]
Tech. Rep. Archives
 Go Top Page Go Previous   [Japanese] / [English] 

Paper Abstract and Keywords
Presentation 2020-12-11 14:00
"Fugaku" Supercomputer Interconnect Failure Prediction Through Deep Learning
Atsushi Miki (Fujitsu) R2020-30
Abstract (in Japanese) (See Japanese page) 
(in English) Supercomputers have large-scale structures to achieve high performance. To complete the benchmark, each unit in the system requires extremely high reliability which is equivalent to several decades of stability. For realizing high reliability, the maintenance impact has to be minimized by preparing replacement units in advance to avoid a long period of system-down. However, it was difficult to predict the number of failures conventionally. Therefore, excessive preparation of the replacement units was necessary to ensure enough safety margin, while a much fewer number of replacement units are used in fact. Optimizing the number of replacement units is an essential issue to reduce maintenance costs. Hence, failure prediction through a statistical analysis of features of system failures is necessary. Supercomputers that equip a lot of log management functions have advantages in the machine learning field compared with other systems. In this paper, a failure prediction technology constructed with deep learning is proposed to minimize the number of unnecessary replacement units. The
evaluation results show the high accuracy of the model.
Keyword (in Japanese) (See Japanese page) 
(in English) deep learning / failure prediction / / / / / /  
Reference Info. IEICE Tech. Rep., vol. 120, no. 286, R2020-30, pp. 1-6, Dec. 2020.
Paper # R2020-30 
Date of Issue 2020-12-04 (R) 
ISSN Print edition: ISSN 0913-5685  Online edition: ISSN 2432-6380
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034)
Download PDF R2020-30

Conference Information
Committee R  
Conference Date 2020-12-11 - 2020-12-11 
Place (in Japanese) (See Japanese page) 
Place (in English) Online 
Topics (in Japanese) (See Japanese page) 
Topics (in English) Reliability International Standard, Maintainability, Reliability General 
Paper Information
Registration To R 
Conference Code 2020-12-R 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) "Fugaku" Supercomputer Interconnect Failure Prediction Through Deep Learning 
Sub Title (in English)  
Keyword(1) deep learning  
Keyword(2) failure prediction  
1st Author's Name Atsushi Miki  
1st Author's Affiliation Fujitsu Limited (Fujitsu)
2nd Author's Name  
2nd Author's Affiliation ()
3rd Author's Name  
3rd Author's Affiliation ()
4th Author's Name  
4th Author's Affiliation ()
5th Author's Name  
5th Author's Affiliation ()
6th Author's Name  
6th Author's Affiliation ()
7th Author's Name  
7th Author's Affiliation ()
8th Author's Name  
8th Author's Affiliation ()
9th Author's Name  
9th Author's Affiliation ()
10th Author's Name  
10th Author's Affiliation ()
11th Author's Name  
11th Author's Affiliation ()
12th Author's Name  
12th Author's Affiliation ()
13th Author's Name  
13th Author's Affiliation ()
14th Author's Name  
14th Author's Affiliation ()
15th Author's Name  
15th Author's Affiliation ()
16th Author's Name  
16th Author's Affiliation ()
17th Author's Name  
17th Author's Affiliation ()
18th Author's Name  
18th Author's Affiliation ()
19th Author's Name  
19th Author's Affiliation ()
20th Author's Name  
20th Author's Affiliation ()
Date Time 2020-12-11 14:00:00 
Presentation Time 25 
Registration for R 
Paper # IEICE-R2020-30 
Volume (vol) IEICE-120 
Number (no) no.286 
Page pp.1-6 
#Pages IEICE-6 
Date of Issue IEICE-R-2020-12-04 

[Return to Top Page]

[Return to IEICE Web Page]

The Institute of Electronics, Information and Communication Engineers (IEICE), Japan