Framework for the hybrid parallelisation of simulation codes

R. P. Mundani; M. Ljucović; E. Rank

Framework for the hybrid parallelisation of simulation codes

R. P. Mundani, M. Ljucović, E. Rank

Engineering and Design

Publikation: Beitrag in Buch/Bericht/Konferenzband › Konferenzbeitrag › Begutachtung

Abstract

Writing efficient hybrid parallel code is tedious, error-prone, and requires good knowledge of both parallel programming and multithreading such as MPI and OpenMP, resp. Therefore, we present a framework which is based on a job model that allows the user to incorporate his sequential code with manageable effort and code modifications in order to be executed in parallel on clusters or supercomputers built from modern multi-core CPUs. The primary application domain of this framework are simulation codes from engineering disciplines as those are in many cases still sequential and due to their memory and runtime demands prominent candidates for parallelisation.

Originalsprache	Englisch
Titel	Proceedings of the 2nd International Conference on Parallel, Distributed, Grid and Cloud Computing for Engineering, PARENG 2011
Herausgeber (Verlag)	Civil-Comp Press
Band	95
ISBN (Print)	9781905088423
Publikationsstatus	Veröffentlicht - 2011
Veranstaltung	2nd International Conference on Parallel, Distributed, Grid and Cloud Computing for Engineering, PARENG 2011 - Ajaccio, Corsica, Frankreich Dauer: 12 Apr. 2011 → 15 Apr. 2011

Konferenz

Konferenz	2nd International Conference on Parallel, Distributed, Grid and Cloud Computing for Engineering, PARENG 2011
Land/Gebiet	Frankreich
Ort	Ajaccio, Corsica
Zeitraum	12/04/11 → 15/04/11

Andere Dateien und Links

Verknüpfung zur Publikation in Scopus

Dieses zitieren

@inproceedings{31939b0350424495befb2896d8dfa7f1,

title = "Framework for the hybrid parallelisation of simulation codes",

abstract = "Writing efficient hybrid parallel code is tedious, error-prone, and requires good knowledge of both parallel programming and multithreading such as MPI and OpenMP, resp. Therefore, we present a framework which is based on a job model that allows the user to incorporate his sequential code with manageable effort and code modifications in order to be executed in parallel on clusters or supercomputers built from modern multi-core CPUs. The primary application domain of this framework are simulation codes from engineering disciplines as those are in many cases still sequential and due to their memory and runtime demands prominent candidates for parallelisation.",

keywords = "Framework, Hybrid, MPI, Openmp, Parallelisation, Simulation",

author = "Mundani, {R. P.} and M. Ljucovi{\'c} and E. Rank",

year = "2011",

language = "English",

isbn = "9781905088423",

volume = "95",

booktitle = "Proceedings of the 2nd International Conference on Parallel, Distributed, Grid and Cloud Computing for Engineering, PARENG 2011",

publisher = "Civil-Comp Press",

note = "2nd International Conference on Parallel, Distributed, Grid and Cloud Computing for Engineering, PARENG 2011 ; Conference date: 12-04-2011 Through 15-04-2011",

}

Mundani, RP, Ljucović, M & Rank, E 2011, Framework for the hybrid parallelisation of simulation codes. in Proceedings of the 2nd International Conference on Parallel, Distributed, Grid and Cloud Computing for Engineering, PARENG 2011. Bd. 95, Civil-Comp Press, 2nd International Conference on Parallel, Distributed, Grid and Cloud Computing for Engineering, PARENG 2011, Ajaccio, Corsica, Frankreich, 12/04/11.

TY - GEN

T1 - Framework for the hybrid parallelisation of simulation codes

AU - Mundani, R. P.

AU - Ljucović, M.

AU - Rank, E.

PY - 2011

Y1 - 2011

N2 - Writing efficient hybrid parallel code is tedious, error-prone, and requires good knowledge of both parallel programming and multithreading such as MPI and OpenMP, resp. Therefore, we present a framework which is based on a job model that allows the user to incorporate his sequential code with manageable effort and code modifications in order to be executed in parallel on clusters or supercomputers built from modern multi-core CPUs. The primary application domain of this framework are simulation codes from engineering disciplines as those are in many cases still sequential and due to their memory and runtime demands prominent candidates for parallelisation.

AB - Writing efficient hybrid parallel code is tedious, error-prone, and requires good knowledge of both parallel programming and multithreading such as MPI and OpenMP, resp. Therefore, we present a framework which is based on a job model that allows the user to incorporate his sequential code with manageable effort and code modifications in order to be executed in parallel on clusters or supercomputers built from modern multi-core CPUs. The primary application domain of this framework are simulation codes from engineering disciplines as those are in many cases still sequential and due to their memory and runtime demands prominent candidates for parallelisation.

KW - Framework

KW - Hybrid

KW - MPI

KW - Openmp

KW - Parallelisation

KW - Simulation

UR - http://www.scopus.com/inward/record.url?scp=84894210951&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84894210951

SN - 9781905088423

VL - 95

BT - Proceedings of the 2nd International Conference on Parallel, Distributed, Grid and Cloud Computing for Engineering, PARENG 2011

PB - Civil-Comp Press

T2 - 2nd International Conference on Parallel, Distributed, Grid and Cloud Computing for Engineering, PARENG 2011

Y2 - 12 April 2011 through 15 April 2011

ER -

Framework for the hybrid parallelisation of simulation codes

Abstract

Konferenz

Andere Dateien und Links

Fingerprint

Dieses zitieren