On the Energy Efficiency and Performance of Irregular Application Executions on Multicore, NUMA and Manycore Platforms - Laboratoire d'Informatique de Grenoble Accéder directement au contenu
Article Dans Une Revue Journal of Parallel and Distributed Computing Année : 2015

On the Energy Efficiency and Performance of Irregular Application Executions on Multicore, NUMA and Manycore Platforms

Résumé

Until the last decade, performance of HPC architectures has been almost exclusively quantifiedby their processing power. However, energy efficiency is being recently considered as importantas raw performance and has become a critical aspect to the development of scalablesystems. These strict energy constraints guided the development of a new class of so-calledlight-weight manycore processors. This study evaluates the computing and energy performanceof two well-known irregular NP-hard problems — the Traveling-Salesman Problem (TSP) andK-Means clustering—and a numerical seismic wave propagation simulation kernel—Ondes3D—on multicore, NUMA, and manycore platforms. First, we concentrate on the nontrivial task ofadapting these applications to a manycore, specifically the novel MPPA-256 manycore processor.Then, we analyze their performance and energy consumption on those di↵erent machines.Our results show that applications able to fully use the resources of a manycore can have betterperformance and may consume from 3.8x to 13x less energy when compared to low-power andgeneral-purpose multicore processors, respectively
Fichier principal
Vignette du fichier
Francesquini_et_al_JPDC_2014.pdf (1.38 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01092325 , version 1 (08-12-2014)

Licence

Paternité - Pas d'utilisation commerciale

Identifiants

Citer

Emilio Francesquini, Márcio Castro, Pedro Henrique Penna, Fabrice Dupros, Henrique Cota De Freitas, et al.. On the Energy Efficiency and Performance of Irregular Application Executions on Multicore, NUMA and Manycore Platforms. Journal of Parallel and Distributed Computing, 2015, 76, pp. 32-48. ⟨10.1016/j.jpdc.2014.11.002⟩. ⟨hal-01092325⟩
1171 Consultations
667 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More