Resource-driven optimizations for transient-fault detecting superscalar microarchitectures

Jie S. Hu, G. M. Link, Johnsy K. John, Shuai Wang, Sotirios G. Ziavras

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Scopus citations

Abstract

Increasing microprocessor vulnerability to soft errors induced by neutron and alpha particle strikes prevents aggressive scaling and integration of transistors in future technologies if left unaddressed. Previously proposed instruction-level redundant execution, as a means of detecting errors, suffers from a severe performance loss due to the resource shortage caused by the large number of redundant instructions injected into the superscalar core. In this paper, we propose to apply three architectural enhancements, namely 1) floating-point unit sharing (FUS), 2) prioritizing primary instructions (PRI), and 3) early retiring of redundant instructions (ERT), that enable transient-fault detecting redundant execution in superscalar microarchitectures with a much smaller performance penalty, while maintaining the original full coverage of soft errors. In addition, our enhancements are compatible with many other proposed techniques, allowing for further performance improvement.

Original languageEnglish (US)
Title of host publicationAdvances in Computer Systems Architecture - 10th Asia-Pacific Conference, ACSAC 2005, Proceedings
PublisherSpringer Verlag
Pages200-214
Number of pages15
ISBN (Print)3540296433, 9783540296430
DOIs
StatePublished - 2005
Event10th Asia-Pacific Conference on Advances in Computer Systems Architecture, ACSAC 2005 - Singapore, Singapore
Duration: Oct 24 2005Oct 26 2005

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3740 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other10th Asia-Pacific Conference on Advances in Computer Systems Architecture, ACSAC 2005
Country/TerritorySingapore
CitySingapore
Period10/24/0510/26/05

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint

Dive into the research topics of 'Resource-driven optimizations for transient-fault detecting superscalar microarchitectures'. Together they form a unique fingerprint.

Cite this