Multibillion-atom molecular dynamics simulation: Design considerations for vector-parallel processing

D. C. Rapaport

doi:10.1016/j.cpc.2005.11.008

Multibillion-atom molecular dynamics simulation: Design considerations for vector-parallel processing

D. C. Rapaport

Department of Physics

Oak Ridge National Laboratory

Research output: Contribution to journal › Article › peer-review

14 Scopus citations

Abstract

Progress in adapting molecular dynamics algorithms for systems with short-range interactions to utilize the features of modern supercomputers is described. Efficient utilization of the latest generation of processor architectures requires algorithms that can be both vectorized and parallelized. The approach adopted for vectorization involves combining the layer and neighbor-list methods, while parallelization employs spatial subdivision with explicit communication. The techniques presented here have been used in performance tests on the Cray X1 vector-parallel supercomputer with systems containing over 12 billion atoms.

Original language	English
Pages (from-to)	521-529
Number of pages	9
Journal	Computer Physics Communications
Volume	174
Issue number	7
DOIs	https://doi.org/10.1016/j.cpc.2005.11.008
State	Published - 1 Apr 2006

Bibliographical note

Funding Information:
J. Barhen, Y. Braiman, P. Cummings, M. Fahey, and J. Nichols are thanked for helpful discussion. This research was sponsored by the Laboratory Directed Research and Development Program of Oak Ridge National Laboratory (ORNL) and used the high-performance computing resources of the Center for Computational Sciences at ORNL which is supported by the Office of Science of the U.S. Department of Energy (DOE). ORNL is managed by UT-Battelle, LLC for DOE under Contract No. DE-AC05-00OR22725.

Funding

J. Barhen, Y. Braiman, P. Cummings, M. Fahey, and J. Nichols are thanked for helpful discussion. This research was sponsored by the Laboratory Directed Research and Development Program of Oak Ridge National Laboratory (ORNL) and used the high-performance computing resources of the Center for Computational Sciences at ORNL which is supported by the Office of Science of the U.S. Department of Energy (DOE). ORNL is managed by UT-Battelle, LLC for DOE under Contract No. DE-AC05-00OR22725.

Funders	Funder number
U.S. Department of Energy	DE-AC05-00OR22725
Office of Science
Oak Ridge National Laboratory

Keywords

Algorithm
Molecular dynamics simulation
Parallel processing
Performance evaluation
Supercomputer
Vector processing

Access to Document

10.1016/j.cpc.2005.11.008

Cite this

@article{94d93a6638374751bf6a8181ea0ad6e9,

title = "Multibillion-atom molecular dynamics simulation: Design considerations for vector-parallel processing",

abstract = "Progress in adapting molecular dynamics algorithms for systems with short-range interactions to utilize the features of modern supercomputers is described. Efficient utilization of the latest generation of processor architectures requires algorithms that can be both vectorized and parallelized. The approach adopted for vectorization involves combining the layer and neighbor-list methods, while parallelization employs spatial subdivision with explicit communication. The techniques presented here have been used in performance tests on the Cray X1 vector-parallel supercomputer with systems containing over 12 billion atoms.",

keywords = "Algorithm, Molecular dynamics simulation, Parallel processing, Performance evaluation, Supercomputer, Vector processing",

author = "Rapaport, {D. C.}",

note = "Funding Information: J. Barhen, Y. Braiman, P. Cummings, M. Fahey, and J. Nichols are thanked for helpful discussion. This research was sponsored by the Laboratory Directed Research and Development Program of Oak Ridge National Laboratory (ORNL) and used the high-performance computing resources of the Center for Computational Sciences at ORNL which is supported by the Office of Science of the U.S. Department of Energy (DOE). ORNL is managed by UT-Battelle, LLC for DOE under Contract No. DE-AC05-00OR22725.",

year = "2006",

month = apr,

day = "1",

doi = "10.1016/j.cpc.2005.11.008",

language = "אנגלית",

volume = "174",

pages = "521--529",

journal = "Computer Physics Communications",

issn = "0010-4655",

publisher = "Elsevier B.V.",

number = "7",

}

TY - JOUR

T1 - Multibillion-atom molecular dynamics simulation

T2 - Design considerations for vector-parallel processing

AU - Rapaport, D. C.

N1 - Funding Information: J. Barhen, Y. Braiman, P. Cummings, M. Fahey, and J. Nichols are thanked for helpful discussion. This research was sponsored by the Laboratory Directed Research and Development Program of Oak Ridge National Laboratory (ORNL) and used the high-performance computing resources of the Center for Computational Sciences at ORNL which is supported by the Office of Science of the U.S. Department of Energy (DOE). ORNL is managed by UT-Battelle, LLC for DOE under Contract No. DE-AC05-00OR22725.

PY - 2006/4/1

Y1 - 2006/4/1

N2 - Progress in adapting molecular dynamics algorithms for systems with short-range interactions to utilize the features of modern supercomputers is described. Efficient utilization of the latest generation of processor architectures requires algorithms that can be both vectorized and parallelized. The approach adopted for vectorization involves combining the layer and neighbor-list methods, while parallelization employs spatial subdivision with explicit communication. The techniques presented here have been used in performance tests on the Cray X1 vector-parallel supercomputer with systems containing over 12 billion atoms.

AB - Progress in adapting molecular dynamics algorithms for systems with short-range interactions to utilize the features of modern supercomputers is described. Efficient utilization of the latest generation of processor architectures requires algorithms that can be both vectorized and parallelized. The approach adopted for vectorization involves combining the layer and neighbor-list methods, while parallelization employs spatial subdivision with explicit communication. The techniques presented here have been used in performance tests on the Cray X1 vector-parallel supercomputer with systems containing over 12 billion atoms.

KW - Algorithm

KW - Molecular dynamics simulation

KW - Parallel processing

KW - Performance evaluation

KW - Supercomputer

KW - Vector processing

UR - http://www.scopus.com/inward/record.url?scp=33644780817&partnerID=8YFLogxK

U2 - 10.1016/j.cpc.2005.11.008

DO - 10.1016/j.cpc.2005.11.008

M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???

AN - SCOPUS:33644780817

SN - 0010-4655

VL - 174

SP - 521

EP - 529

JO - Computer Physics Communications

JF - Computer Physics Communications

IS - 7

ER -

Multibillion-atom molecular dynamics simulation: Design considerations for vector-parallel processing

Abstract

Bibliographical note

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this