%0 Conference Proceedings
%T An Adaptive Low-Overhead Mechanism for Dependable General-Purpose Many-Core Processors
%+ National University of Defense Technology [China]
%A Jia, Wentao
%A Li, Rui
%A Zhang, Chunyan
%Z Part 2: Asian Conference on Availability, Reliability and Security (AsiaARES)
%< avec comité de lecture
%( Lecture Notes in Computer Science
%B 1st International Conference on Information and Communication Technology (ICT-EurAsia)
%C Yogyakarta, Indonesia
%Y David Hutchison
%Y Takeo Kanade
%Y Madhu Sudan
%Y Demetri Terzopoulos
%Y Doug Tygar
%Y Moshe Y. Vardi
%Y Gerhard Weikum
%Y Khabib Mustofa
%Y Erich J. Neuhold
%Y A Min Tjoa
%Y Edgar Weippl
%Y Ilsun You
%Y Josef Kittler
%Y Jon M. Kleinberg
%Y Friedemann Mattern
%Y John C. Mitchell
%Y Moni Naor
%Y Oscar Nierstrasz
%Y C. Pandu Rangan
%Y Bernhard Steffen
%I Springer
%3 Information and Communicatiaon Technology
%V LNCS-7804
%P 337-342
%8 2013-03-25
%D 2013
%R 10.1007/978-3-642-36818-9_37
%K Many-Core
%K Redundant Execution
%K Adaptive Dependable
%K Low-Overhead
%Z Computer Science [cs]
%Z Humanities and Social Sciences/Library and information sciencesConference papers
%X Future many-core processors may contain more than 1000 cores on single die. However, continued scaling of silicon fabrication technology exposes chip orders of such magnitude to a higher vulnerability to errors. A low-overhead and adaptive fault-tolerance mechanism is desired for general-purpose many-core processors. We propose high-level adaptive redundancy (HLAR), which possesses several unique properties. First, the technique employs selective redundancy based application assistance and dynamically cores schedule. Second, the method requires minimal overhead when the mechanism is disabled. Third, it expands the local memory within the replication sphere, which heightens the replication level and simplifies the redundancy mechanism. Finally, it decreases bandwidth through various compression methods, thus effectively balancing reliability, performance, and power. Experimental results show a remarkably low overhead while covering 99.999% errors with only 0.25% more networks-on-chip traffic.
%G English
%Z TC 5
%Z TC 8
%2 https://inria.hal.science/hal-01480191/document
%2 https://inria.hal.science/hal-01480191/file/978-3-642-36818-9_37_Chapter.pdf
%L hal-01480191
%U https://inria.hal.science/hal-01480191
%~ SHS
%~ IFIP-LNCS
%~ IFIP
%~ IFIP-TC
%~ IFIP-TC5
%~ IFIP-TC8
%~ IFIP-ICT-EURASIA
%~ IFIP-LNCS-7804