Algorithm-based fault tolerance for dense matrix factorizations

Author:

Du Peng1,Bouteiller Aurelien1,Bosilca George1,Herault Thomas1,Dongarra Jack1

Affiliation:

1. University of Tennessee, Knoxville, Knoxville, TN, USA

Abstract

Dense matrix factorizations, such as LU, Cholesky and QR, are widely used for scientific applications that require solving systems of linear equations, eigenvalues and linear least squares problems. Such computations are normally carried out on supercomputers, whose ever-growing scale induces a fast decline of the Mean Time To Failure (MTTF). This paper proposes a new hybrid approach, based on Algorithm-Based Fault Tolerance (ABFT), to help matrix factorizations algorithms survive fail-stop failures. We consider extreme conditions, such as the absence of any reliable component and the possibility of loosing both data and checksum from a single failure. We will present a generic solution for protecting the right factor, where the updates are applied, of all above mentioned factorizations. For the left factor, where the panel has been applied, we propose a scalable checkpointing algorithm. This algorithm features high degree of checkpointing parallelism and cooperatively utilizes the checksum storage leftover from the right factor protection. The fault-tolerant algorithms derived from this hybrid solution is applicable to a wide range of dense matrix factorizations, with minor modifications. Theoretical analysis shows that the fault tolerance overhead sharply decreases with the scaling in the number of computing units and the problem size. Experimental results of LU and QR factorization on the Kraken (Cray XT5) supercomputer validate the theoretical evaluation and confirm negligible overhead, with- and without-errors.

Publisher

Association for Computing Machinery (ACM)

Subject

Computer Graphics and Computer-Aided Design,Software

Reference22 articles.

1. Fault tolerance for extreme-scale computing workshop report 2009. Fault tolerance for extreme-scale computing workshop report 2009.

2. http://www.top500.org/ 2011. http://www.top500.org/ 2011.

3. ScaLAPACK Users' Guide

4. Algorithm-based fault tolerance applied to high performance computing

Cited by 70 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

1. Rollback-Free Recovery for a High Performance Dense Linear Solver With Reduced Memory Footprint;IEEE Transactions on Parallel and Distributed Systems;2024-07

2. Extending the Legio Resilience Framework to Handle Critical Process Failures in MPI;2024 32nd Euromicro International Conference on Parallel, Distributed and Network-Based Processing (PDP);2024-03-20

3. An overview of the Legio fault resilience framework for MPI applications;Procedia Computer Science;2024

4. Automatic Algorithm-Based Fault Tolerance (AABFT) of Stencil Computations;2023 32nd International Conference on Parallel Architectures and Compilation Techniques (PACT);2023-10-21

5. The Legio Fault Resilience Framework;Proceedings of the 20th ACM International Conference on Computing Frontiers;2023-05-09

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3