An Adaptive Low-Overhead Mechanism for Dependable General-Purpose Many-Core Processors - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2013

An Adaptive Low-Overhead Mechanism for Dependable General-Purpose Many-Core Processors

Wentao Jia
  • Function : Author
  • PersonId : 1003088
Rui Li
  • Function : Author
  • PersonId : 1003089

Abstract

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.
Fichier principal
Vignette du fichier
978-3-642-36818-9_37_Chapter.pdf (219.83 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01480191 , version 1 (01-03-2017)

Licence

Attribution

Identifiers

Cite

Wentao Jia, Rui Li, Chunyan Zhang. An Adaptive Low-Overhead Mechanism for Dependable General-Purpose Many-Core Processors. 1st International Conference on Information and Communication Technology (ICT-EurAsia), Mar 2013, Yogyakarta, Indonesia. pp.337-342, ⟨10.1007/978-3-642-36818-9_37⟩. ⟨hal-01480191⟩
41 View
80 Download

Altmetric

Share

Gmail Facebook X LinkedIn More