A Backward/Forward Recovery Approach for the Preconditioned Conjugate Gradient Algorithm

Abstract : Several recent papers have introduced a periodic verification mechanism to detect silent errors in iterative solvers. Chen [PPoPP'13, pp. 167--176] has shown how to combine such a verification mechanism (a stability test checking the orthogonality of two vectors and recomputing the residual) with checkpointing: the idea is to verify every $d$ iterations, and to checkpoint every $c \times d$ iterations. When a silent error is detected by the verification mechanism, one can rollback to and re-execute from the last checkpoint. In this paper, we also propose to combine checkpointing and verification, but we use algorithm-based fault tolerance (ABFT) rather than stability tests. ABFT can be used for error detection, but also for error detection and correction, allowing a forward recovery (and no rollback nor re-execution) when a single error is detected. We introduce an abstract performance model to compute the performance of all schemes, and we instantiate it using the preconditioned conjugate gradient algorithm. Finally, we validate our new approach through a set of simulations.
Type de document :
Rapport
[Research Report] RR-8826, ENS Lyon, CNRS & INRIA. 2015
Liste complète des métadonnées

Littérature citée [44 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-01242327
Contributeur : Equipe Roma <>
Soumis le : vendredi 11 décembre 2015 - 19:15:13
Dernière modification le : vendredi 20 avril 2018 - 15:44:27
Document(s) archivé(s) le : samedi 29 avril 2017 - 11:47:55

Fichier

RR-8826.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01242327, version 1

Collections

Citation

Massimiliano Fasi, Julien Langou, Yves Robert, Bora Uçar. A Backward/Forward Recovery Approach for the Preconditioned Conjugate Gradient Algorithm. [Research Report] RR-8826, ENS Lyon, CNRS & INRIA. 2015. 〈hal-01242327〉

Partager

Métriques

Consultations de la notice

325

Téléchargements de fichiers

84