Idle waves in high-performance computing

Abstract : The vast majority of parallel scientific applications distributes computation among processes that are in a busy state when computing and in an idle state when waiting for information from other processes. We identify the propagation of idle waves through processes in scientific applications with a local information exchange between the two processes. Idle waves are nondispersive and have a phase velocity inversely proportional to the average busy time. The physical mechanism enabling the propagation of idle waves is the local synchronization between two processes due to remote data dependency. This study provides a description of the large number of processes in parallel scientific applications as a continuous medium. This work also is a step towards an understanding of how localized idle periods can affect remote processes, leading to the degradation of global performance in parallel scientific applications.
Type de document :
Article dans une revue
Physical Review E : Statistical, Nonlinear, and Soft Matter Physics, American Physical Society, 2015, 91, pp.13306 - 13306. 〈10.1103/PhysRevE.91.013306〉
Liste complète des métadonnées

Littérature citée [5 références]  Voir  Masquer  Télécharger

https://hal-insu.archives-ouvertes.fr/insu-01397212
Contributeur : Nathalie Pothier <>
Soumis le : mardi 15 novembre 2016 - 15:42:51
Dernière modification le : jeudi 7 février 2019 - 17:05:52
Document(s) archivé(s) le : jeudi 16 mars 2017 - 18:43:29

Fichier

PhysRevE.91.pdf
Fichiers éditeurs autorisés sur une archive ouverte

Identifiants

Collections

Citation

Stefano Markidis, Juris Vencels, Ivy Peng, Dana Akhmetova, Erwin Laure, et al.. Idle waves in high-performance computing. Physical Review E : Statistical, Nonlinear, and Soft Matter Physics, American Physical Society, 2015, 91, pp.13306 - 13306. 〈10.1103/PhysRevE.91.013306〉. 〈insu-01397212〉

Partager

Métriques

Consultations de la notice

99

Téléchargements de fichiers

95