CCGrid 2005. IEEE International Symposium on Cluster Computing and the Grid, 2005. 2005
DOI: 10.1109/ccgrid.2005.1558650
|View full text |Cite
|
Sign up to set email alerts
|

A new dynamical domain decomposition method for parallel molecular dynamics simulation

Abstract: A new material particle dynamical domain decomposition method MPD3 has been developed. The method is suitable for a large scale parallel molecular dynamic simulation on a heterogeneous computing net. Performance of the MPD3 algorithm is tested in various computing environments, such as PC clusters, super computer clusters, and Grid. It is shown that the MPD3 algorithm is highly adaptivefor both computer clusters and Grid computing environments, even ifotherprograms are running on the same computer environment.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
22
0

Year Published

2007
2007
2020
2020

Publication Types

Select...
5
2
1

Relationship

1
7

Authors

Journals

citations
Cited by 24 publications
(22 citation statements)
references
References 5 publications
0
22
0
Order By: Relevance
“…In this research for hydrodynamic simulations the Lagrangian mesh-free SPH method [6,7] is used. Atomistic simulations are performed with an autobalancing parallel MD code (MD-MPD 3 [8]) based on dynamical domain decomposition of samples with highly-nonuniform density distributions into dynamical Voronoi polyhedrons.…”
Section: Introductionmentioning
confidence: 99%
“…In this research for hydrodynamic simulations the Lagrangian mesh-free SPH method [6,7] is used. Atomistic simulations are performed with an autobalancing parallel MD code (MD-MPD 3 [8]) based on dynamical domain decomposition of samples with highly-nonuniform density distributions into dynamical Voronoi polyhedrons.…”
Section: Introductionmentioning
confidence: 99%
“…The grey continuous curve separates non-ejecting ( the blue circles) and ejecting cases ( the red circles). We see that the simulations rather satisfactory correspond to the experiments Main new achievement used in the paper here (relative to the previous work [ 22]) is addition of the Monte-Carlo (MC) algorithm to the molecular dynamics (MD) algorithm [ 27]. The sense of the MC modeling is the following.…”
Section: Figmentioning
confidence: 91%
“…Zhakhovskii et al [1] introduced a decomposition scheme that divides the simulation domain into Voronoi regions, whose size and shape are varied during the simulation to alleviate the imbalances in processor runtimes. Even though Voronoi decomposition is an elegant way of subdivision, it comes with the drawback of dealing with the complicated nature of 3D Voronoi regions, which have to be updated gradually.…”
Section: Related Workmentioning
confidence: 99%