Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ HAL-ENS-LYONarrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
INRIA2
External research report . 2017
Data sources: INRIA2
https://doi.org/10.1109/icpp.2...
Article . 2017 . Peer-reviewed
Data sources: Crossref
DBLP
Conference object
Data sources: DBLP
versions View all 6 versions
addClaim

Resilience for Stencil Computations with Latent Errors

Authors: Fang, Aiman; Cavelan, Aurélien; Robert, Yves; Chien, Andrew;

Resilience for Stencil Computations with Latent Errors

Abstract

Les projections et mesures pour les systèmes exascale ($10^9$ coeurs) suggèrent une augmentation très importante du taux d’erreur. Une telle augmentation fait de la résilience un sujet critique, et risque d’aggraverl’impact des erreurs qui “s’échappent”, corrompant silencieusement la mémoire. Ces erreurs sont souvent détectées par des tests logiciels au niveau de l’application, mais avec une latence de détection importante, et sont donc connues sous le nom d’erreurs latentes. Nous explorons une approche appeléeapplication-based-focus-recovery, ou ABFR, afin de relancer l’exécution efficacement, suit à une erreur. En particulier, nous présentons une étudede cas pour les applications de type stencil, montrant comment ABFR concentre les calculs de récupération où ils sont nécessaire, utilisant des tests et des élagages intelligents pour réduire les calculs de récupération, et permettre le recouvrement avec les calculs de l’application. Nous analysons et caractérisons l’approche ABFR pour les applications de type stencil, créant un modèle de performance paramétré par le taux d’erreur et l’interval de détection (la latence). Nous comparons les projections du modèle aux résultats expérimentaux avec l’application stencil Chombo, validant le modèle et montrant que ABFR permet d’obtenir une réduction significative du coût derécupération (jusqu’à 400x) et de la latence (jusqu’à 4x). De telles réductions de coût permettent de passer à l’échelle avec des taux d’erreurs latentes élevés.

Projections and measurements of error rates in near-exascale and exascale systems suggest a dramatic growth, due to extreme scale ($10^9$ cores), concurrency, software complexity, and deep submicron transistor scaling.Such a growth makes resilience a critical concern, and may increase the incidence of errors that ``escape'', silently corrupting application state. Such errors can often be revealed by application software tests but with long latencies, and thus are known as {\it latent errors}. We explore how to efficiently recover from latent errors, with an approach called application-based focused recovery (ABFR). Specifically wepresent a case study of stencil computations, a widely useful computational structure, showing how ABFR focuses recovery effortwhere needed, using intelligent testing and pruning to reduce recovery effort, and enables recovery effort to be overlapped with application computation. We analyze and characterize the ABFR approach on stencils, creating a performance model parameterized by error rate and detection interval (latency).We compare projections from the model to experimental resultswith the Chombo stencil application, validating the model and showing that ABFR on stencil can achieve a significant reductions in error recovery cost (up to 400x) and recovery latency (up to 4x).Such reductions enable efficient execution at scale with high latent error rates.

Country
France
Keywords

stencil, optimal pattern, fail-stop errors, ABFR, multi-level checkpointing, [INFO] Computer Science [cs], erreurs latentes, résilience

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    7
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Top 10%
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
7
Average
Average
Top 10%
Green