logo
科言猫学术研究的AI总结
language
手机端手机端
  • Phone Feature微信小程序
简体中文简体中文
notice消息通知
博客博客
delete
arrow
返回

Fault-tolerant numerical iterative algorithms at scale

delete2025-10-03
delete0
PRE
AI
A
Alix Tremodeux
A
Anne Benoît
E
Emmanuel Agullo
T
Thomas Hérault
L
Luc Giraud
Y
Yves Robert
DOI：10.1177/10943420251379675delete
delete原文链接
delete原文求助
delete分享
delete收藏
摘要

摘要

En 中文
本研究探讨了如何保护数值迭代算法免受所有可能在大规模系统中发生的错误类型的影响：失效停止错误（又称故障）和静默错误，这些错误既表现为计算错误，也表现为内存位翻转。我们结合了多种技术：用于检测计算错误的探测器、用于检测内存错误的校验和，以及用于应对故障的检查点/重启机制。目标是使算法每次迭代的预期时间最小化。我们设计了一种分层模式，将所有这些容错机制进行组合和交错处理，并确定了实现该目标的最佳周期性模式。我们将这些结果应用于预条件共轭梯度（PCG）算法的性能分析：我们报告了多种场景，在这些场景中，最佳模式显著降低了错误缓解带来的开销。

期刊

T
The International Journal of High Performance Computing Applications
IF:
0
论文数:
49
被引数:
0

机构

I
inria centre at the university of bordeaux, talence, france
学者数:
3
论文数: 1
被引数: 0
L
lip & inria, ens lyon, lyon, france
学者数:
3
论文数: 1
被引数: 0
引用论文

引用论文

A Large-Scale Study of Failures on Petascale Supercomputers
errJournal of Computer Science and Technology
IF0
err2018-01-26
err0
PREAI
errRui-Tao Liu; Zuo-Ning Chen
err分享
err收藏
Unprotected Computing: A Large-Scale Study of DRAM Raw Error Rate on a Supercomputer
errSC16: International Conference for High Performance Computing, Networking, Storage and Analysis
IF0
err2016-11-01
err0
errOAAI
errLeonardo Bautista-Gomez; Ferad Zyulkyarov; Osman Unsal; Simon McIntosh-Smith
err分享
err收藏
Hydrostatic, quasi-hydrostatic, and nonhydrostatic ocean modeling
errJOURNAL OF GEOPHYSICAL RESEARCH-OCEANS
IF3.4
err1997-03-15
err1.2K
errOAAI
errMarshall, J; Hill, C; Perelman, L; Adcroft, A
err分享
err收藏
Design, Modeling, and Evaluation of a Scalable Multi-level Checkpointing System
err2010 ACM/IEEE International Conference for High Performance Computing, Networking, Storage and Analysis
IF0
err2010-11-01
err0
PREAI
errAdam Moody; Greg Bronevetsky; Kathryn Mohror; Bronis R. de Supinski
err分享
err收藏
Cosmic rays don't strike twice
errACM SIGPLAN Notices
IF0
err2012-03-03
err0
PREAI
errAndy A. Hwang; Ioan A. Stefanovici; Bianca Schroeder
err分享
err收藏
Detection and correction of silent errors in the conjugate gradient algorithm
errNumerical Algorithms
IF0
err2022-07-29
err0
PREAI
errGérard Meurant
err分享
err收藏
Understanding Soft Error Resiliency of Blue Gene/Q Compute Chip through Hardware Proton Irradiation and Software Fault Injection
errSC14: International Conference for High Performance Computing, Networking, Storage and Analysis
IF0
err2014-11-01
err0
PREAI
errChen-Yong Cher; Meeta S. Gupta; Pradip Bose; K. Paul Muller
err分享
err收藏
A backward/forward recovery approach for the preconditioned conjugate gradient method
errnature computational science
IF3.7
err2016-11-01
err13
errOAAI
errFasi, Massimiliano; Langou, Julien; Robert, Yves; Ucar, Bora
err分享
err收藏
Frontier: Exploring Exascale
errProceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis
IF0
err2023-11-11
err0
errOAAI
errScott Atchley; Christopher Zimmer; John Lange; David Bernholdt; Veronica Melesse Vergara; Thomas Beck; Michael Brim; Reuben Budiardja; Sunita Chandrasekaran; Markus Eisenbach; Thomas Evans; Matthew Ezell; Nicholas Frontiere; Antigoni Georgiadou; Joe Glenski; Philipp Grete; Steven Hamilton; John Holmen; Axel Huebl; Daniel Jacobson; Wayne Joubert; Kim Mcmahon; Elia Merzari; Stan Moore; Andrew Myers; Stephen Nichols; Sarp Oral; Thomas Papatheodore; Danny Perez; David M. Rogers; Evan Schneider; Jean-Luc Vay; P. K. Yeung
err分享
err收藏
Analyzing the Impact of System Reliability Events on Applications in the Titan Supercomputer
err2018 IEEE/ACM 8th Workshop on Fault Tolerance for HPC at eXtreme Scale (FTXS)
IF0
err2018-11-01
err0
PREAI
errAshraf,Rizwan A.; Engelmann,Christian
err分享
err收藏
学者 查看更多内容