Highly scalable FETI methods in ESPRESO

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F61989100%3A27740%2F15%3A86099018" target="_blank" >RIV/61989100:27740/15:86099018 - isvavai.cz</a>

Nalezeny alternativní kódy

RIV/61989100:27230/15:86099018

Výsledek na webu

<a href="http://dx.doi.org/10.4203/ccp.107.17" target="_blank" >http://dx.doi.org/10.4203/ccp.107.17</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.4203/ccp.107.17" target="_blank" >10.4203/ccp.107.17</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Highly scalable FETI methods in ESPRESO

Popis výsledku v původním jazyce

This paper describes our new hybrid parallelization of the Finite Element Tearing and Interconnecting (FETI) method for the multi-socket and multi-core computer clusters. This is an essential step in our development of the Hybrid FETI solver were a small number of neighboring subdomains is aggregated into clusters and each cluster is processed by a single compute node. In our previous work the FETI solver was implemented using MPI parallelization into our ESPRESO solver. The proposed hybrid implementation provides better utilization of resources of modern HPC machines using advanced shared memory runtime systems such as Cilk++ runtime. Cilk++ is an alternative to OpenMP which is used by the ESPRESO for shared memory parallelization. We have compared the performance of the hybrid parallelization to MPI-only parallelization. The results show that we have reduced both solver runtime and memory utilization. This allows the solver to use a larger number of smaller sub-domains and in order to solve larger problems using a limited number of compute nodes. This feature is essential for users with smaller computer clusters.

Název v anglickém jazyce

Highly scalable FETI methods in ESPRESO

Popis výsledku anglicky

This paper describes our new hybrid parallelization of the Finite Element Tearing and Interconnecting (FETI) method for the multi-socket and multi-core computer clusters. This is an essential step in our development of the Hybrid FETI solver were a small number of neighboring subdomains is aggregated into clusters and each cluster is processed by a single compute node. In our previous work the FETI solver was implemented using MPI parallelization into our ESPRESO solver. The proposed hybrid implementation provides better utilization of resources of modern HPC machines using advanced shared memory runtime systems such as Cilk++ runtime. Cilk++ is an alternative to OpenMP which is used by the ESPRESO for shared memory parallelization. We have compared the performance of the hybrid parallelization to MPI-only parallelization. The results show that we have reduced both solver runtime and memory utilization. This allows the solver to use a larger number of smaller sub-domains and in order to solve larger problems using a limited number of compute nodes. This feature is essential for users with smaller computer clusters.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

Výsledek vznikl pri realizaci vícero projektů. Více informací v záložce Projekty.

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2015

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Civil-Comp proceedings. Volume 107

ISBN

978-1-905088-62-1

ISSN

1759-3433

e-ISSN

neuvedeno

Počet stran výsledku

13

Strana od-do

"nestrankovano"

Název nakladatele

Civil-Comp Press

Místo vydání

Kippen, Stirlingshire

Místo konání akce

Dubrovník

Datum konání akce

24. 3. 2015

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—