Optimized force calculation in molecular dynamics simulations for the intel xeon phi

Nikola Tchipev; Amer Wafai; Colin W. Glass; Wolfgang Eckhardt; Alexander Heinecke; Hans Joachim Bungartz; Philipp Neumann

doi:10.1007/978-3-319-27308-2_62

Optimized force calculation in molecular dynamics simulations for the intel xeon phi

Nikola Tchipev, Amer Wafai, Colin W. Glass, Wolfgang Eckhardt, Alexander Heinecke, Hans Joachim Bungartz, Philipp Neumann

Informatics 5 - Chair of Scientific Computing

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

7 Scopus citations

Abstract

We provide details on the shared-memory parallelization for manycore architectures of the molecular dynamics framework ls1-mardyn, including an optimization of the SIMD vectorization for multi-centered molecules. The novel shared-memory parallelization scheme allows to retain Newton’s third law optimization and exhibits very good scaling on many-core devices such as a full Xeon Phi card running 240 threads. The Xeon Phi can thus be exploited and delivers comparable performance as IvyBridge nodes in our experiments.

Original language	English
Title of host publication	Euro-Par 2015
Subtitle of host publication	Parallel Processing Workshops - Euro-Par 2015 International Workshops, Revised Selected Papers
Editors	Sascha Hunold, Josef Weidendorfer, Domingo Gimenez, Laura Ricci, Stefan Lankes, Alexandru Costan, Ana Lucia Varbanescu, Stephen L. Scott, María Engracia Gómez Requena, Vittorio Scarano, Alexandru Iosup, Michael Alexander
Publisher	Springer Verlag
Pages	774-785
Number of pages	12
ISBN (Print)	9783319273075
DOIs	https://doi.org/10.1007/978-3-319-27308-2_62
State	Published - 2015
Event	International Workshops on Parallel Processing Workshops, Euro-Par 2015 - Vienna, Austria Duration: 24 Aug 2015 → 25 Aug 2015

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9523
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International Workshops on Parallel Processing Workshops, Euro-Par 2015
Country/Territory	Austria
City	Vienna
Period	24/08/15 → 25/08/15

Access to Document

10.1007/978-3-319-27308-2_62

Cite this

Tchipev, N., Wafai, A., Glass, C. W., Eckhardt, W., Heinecke, A., Bungartz, H. J., & Neumann, P. (2015). Optimized force calculation in molecular dynamics simulations for the intel xeon phi. In S. Hunold, J. Weidendorfer, D. Gimenez, L. Ricci, S. Lankes, A. Costan, A. L. Varbanescu, S. L. Scott, M. E. G. Requena, V. Scarano, A. Iosup, & M. Alexander (Eds.), Euro-Par 2015: Parallel Processing Workshops - Euro-Par 2015 International Workshops, Revised Selected Papers (pp. 774-785). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9523). Springer Verlag. https://doi.org/10.1007/978-3-319-27308-2_62

Tchipev, Nikola ; Wafai, Amer ; Glass, Colin W. et al. / Optimized force calculation in molecular dynamics simulations for the intel xeon phi. Euro-Par 2015: Parallel Processing Workshops - Euro-Par 2015 International Workshops, Revised Selected Papers. editor / Sascha Hunold ; Josef Weidendorfer ; Domingo Gimenez ; Laura Ricci ; Stefan Lankes ; Alexandru Costan ; Ana Lucia Varbanescu ; Stephen L. Scott ; María Engracia Gómez Requena ; Vittorio Scarano ; Alexandru Iosup ; Michael Alexander. Springer Verlag, 2015. pp. 774-785 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{e7e1d01d66d84f7081dc48d518b59933,

title = "Optimized force calculation in molecular dynamics simulations for the intel xeon phi",

abstract = "We provide details on the shared-memory parallelization for manycore architectures of the molecular dynamics framework ls1-mardyn, including an optimization of the SIMD vectorization for multi-centered molecules. The novel shared-memory parallelization scheme allows to retain Newton{\textquoteright}s third law optimization and exhibits very good scaling on many-core devices such as a full Xeon Phi card running 240 threads. The Xeon Phi can thus be exploited and delivers comparable performance as IvyBridge nodes in our experiments.",

author = "Nikola Tchipev and Amer Wafai and Glass, \{Colin W.\} and Wolfgang Eckhardt and Alexander Heinecke and Bungartz, \{Hans Joachim\} and Philipp Neumann",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing Switzerland 2015.; International Workshops on Parallel Processing Workshops, Euro-Par 2015 ; Conference date: 24-08-2015 Through 25-08-2015",

year = "2015",

doi = "10.1007/978-3-319-27308-2\_62",

language = "English",

isbn = "9783319273075",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "774--785",

editor = "Sascha Hunold and Josef Weidendorfer and Domingo Gimenez and Laura Ricci and Stefan Lankes and Alexandru Costan and Varbanescu, \{Ana Lucia\} and Scott, \{Stephen L.\} and Requena, \{Mar{\'i}a Engracia G{\'o}mez\} and Vittorio Scarano and Alexandru Iosup and Michael Alexander",

booktitle = "Euro-Par 2015",

}

Tchipev, N, Wafai, A, Glass, CW, Eckhardt, W, Heinecke, A, Bungartz, HJ & Neumann, P 2015, Optimized force calculation in molecular dynamics simulations for the intel xeon phi. in S Hunold, J Weidendorfer, D Gimenez, L Ricci, S Lankes, A Costan, AL Varbanescu, SL Scott, MEG Requena, V Scarano, A Iosup & M Alexander (eds), Euro-Par 2015: Parallel Processing Workshops - Euro-Par 2015 International Workshops, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9523, Springer Verlag, pp. 774-785, International Workshops on Parallel Processing Workshops, Euro-Par 2015, Vienna, Austria, 24/08/15. https://doi.org/10.1007/978-3-319-27308-2_62

Optimized force calculation in molecular dynamics simulations for the intel xeon phi. / Tchipev, Nikola; Wafai, Amer; Glass, Colin W. et al.
Euro-Par 2015: Parallel Processing Workshops - Euro-Par 2015 International Workshops, Revised Selected Papers. ed. / Sascha Hunold; Josef Weidendorfer; Domingo Gimenez; Laura Ricci; Stefan Lankes; Alexandru Costan; Ana Lucia Varbanescu; Stephen L. Scott; María Engracia Gómez Requena; Vittorio Scarano; Alexandru Iosup; Michael Alexander. Springer Verlag, 2015. p. 774-785 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9523).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Optimized force calculation in molecular dynamics simulations for the intel xeon phi

AU - Tchipev, Nikola

AU - Wafai, Amer

AU - Glass, Colin W.

AU - Eckhardt, Wolfgang

AU - Heinecke, Alexander

AU - Bungartz, Hans Joachim

AU - Neumann, Philipp

N1 - Publisher Copyright: © Springer International Publishing Switzerland 2015.

PY - 2015

Y1 - 2015

N2 - We provide details on the shared-memory parallelization for manycore architectures of the molecular dynamics framework ls1-mardyn, including an optimization of the SIMD vectorization for multi-centered molecules. The novel shared-memory parallelization scheme allows to retain Newton’s third law optimization and exhibits very good scaling on many-core devices such as a full Xeon Phi card running 240 threads. The Xeon Phi can thus be exploited and delivers comparable performance as IvyBridge nodes in our experiments.

AB - We provide details on the shared-memory parallelization for manycore architectures of the molecular dynamics framework ls1-mardyn, including an optimization of the SIMD vectorization for multi-centered molecules. The novel shared-memory parallelization scheme allows to retain Newton’s third law optimization and exhibits very good scaling on many-core devices such as a full Xeon Phi card running 240 threads. The Xeon Phi can thus be exploited and delivers comparable performance as IvyBridge nodes in our experiments.

UR - https://www.scopus.com/pages/publications/84951924503

U2 - 10.1007/978-3-319-27308-2_62

DO - 10.1007/978-3-319-27308-2_62

M3 - Conference contribution

AN - SCOPUS:84951924503

SN - 9783319273075

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 774

EP - 785

BT - Euro-Par 2015

A2 - Hunold, Sascha

A2 - Weidendorfer, Josef

A2 - Gimenez, Domingo

A2 - Ricci, Laura

A2 - Lankes, Stefan

A2 - Costan, Alexandru

A2 - Varbanescu, Ana Lucia

A2 - Scott, Stephen L.

A2 - Requena, María Engracia Gómez

A2 - Scarano, Vittorio

A2 - Iosup, Alexandru

A2 - Alexander, Michael

PB - Springer Verlag

T2 - International Workshops on Parallel Processing Workshops, Euro-Par 2015

Y2 - 24 August 2015 through 25 August 2015

ER -

Tchipev N, Wafai A, Glass CW, Eckhardt W, Heinecke A, Bungartz HJ et al. Optimized force calculation in molecular dynamics simulations for the intel xeon phi. In Hunold S, Weidendorfer J, Gimenez D, Ricci L, Lankes S, Costan A, Varbanescu AL, Scott SL, Requena MEG, Scarano V, Iosup A, Alexander M, editors, Euro-Par 2015: Parallel Processing Workshops - Euro-Par 2015 International Workshops, Revised Selected Papers. Springer Verlag. 2015. p. 774-785. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-27308-2_62

Optimized force calculation in molecular dynamics simulations for the intel xeon phi

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this