Optimization of the kinetic activation-relaxation technique, an off-lattice and self-learning kinetic monte-carlo method

Jean François Joly, Laurent Karim Béland, Peter Brommer, Fadwa El-Mellouhi, Normand Mousseau

Research output: Contribution to journalArticle

15 Citations (Scopus)

Abstract

We present two major optimizations for the kinetic Activation-Relaxation Technique (k-ART), an off-lattice self-learning kinetic Monte Carlo (KMC) algorithm with on-the-fly event search THAT has been successfully applied to study a number of semiconducting and metallic systems. K-ART is parallelized in a non-trivial way: A master process uses several worker processes to perform independent event searches for possible events, while all bookkeeping and the actual simulation is performed by the master process. Depending on the complexity of the system studied, the parallelization scales well for tens to more than one hundred processes. For dealing with large systems, we present a near order 1 implementation. Techniques such as Verlet lists, cell decomposition and partial force calculations are implemented, and the CPU time per time step scales sublinearly with the number of particles, providing an efficient use of computational resources.

Original languageEnglish
Article number012007
JournalJournal of Physics: Conference Series
Volume341
Issue number1
DOIs
Publication statusPublished - 22 Mar 2012

Fingerprint

learning
Monte Carlo method
activation
optimization
kinetics
lists
resources
decomposition
cells
simulation

ASJC Scopus subject areas

  • Physics and Astronomy(all)

Cite this

Optimization of the kinetic activation-relaxation technique, an off-lattice and self-learning kinetic monte-carlo method. / Joly, Jean François; Béland, Laurent Karim; Brommer, Peter; El-Mellouhi, Fadwa; Mousseau, Normand.

In: Journal of Physics: Conference Series, Vol. 341, No. 1, 012007, 22.03.2012.

Research output: Contribution to journalArticle

@article{473b5540188d41d2a59a96204a5fcf0f,
title = "Optimization of the kinetic activation-relaxation technique, an off-lattice and self-learning kinetic monte-carlo method",
abstract = "We present two major optimizations for the kinetic Activation-Relaxation Technique (k-ART), an off-lattice self-learning kinetic Monte Carlo (KMC) algorithm with on-the-fly event search THAT has been successfully applied to study a number of semiconducting and metallic systems. K-ART is parallelized in a non-trivial way: A master process uses several worker processes to perform independent event searches for possible events, while all bookkeeping and the actual simulation is performed by the master process. Depending on the complexity of the system studied, the parallelization scales well for tens to more than one hundred processes. For dealing with large systems, we present a near order 1 implementation. Techniques such as Verlet lists, cell decomposition and partial force calculations are implemented, and the CPU time per time step scales sublinearly with the number of particles, providing an efficient use of computational resources.",
author = "Joly, {Jean Fran{\cc}ois} and B{\'e}land, {Laurent Karim} and Peter Brommer and Fadwa El-Mellouhi and Normand Mousseau",
year = "2012",
month = "3",
day = "22",
doi = "10.1088/1742-6596/341/1/012007",
language = "English",
volume = "341",
journal = "Journal of Physics: Conference Series",
issn = "1742-6588",
publisher = "IOP Publishing Ltd.",
number = "1",

}

TY - JOUR

T1 - Optimization of the kinetic activation-relaxation technique, an off-lattice and self-learning kinetic monte-carlo method

AU - Joly, Jean François

AU - Béland, Laurent Karim

AU - Brommer, Peter

AU - El-Mellouhi, Fadwa

AU - Mousseau, Normand

PY - 2012/3/22

Y1 - 2012/3/22

N2 - We present two major optimizations for the kinetic Activation-Relaxation Technique (k-ART), an off-lattice self-learning kinetic Monte Carlo (KMC) algorithm with on-the-fly event search THAT has been successfully applied to study a number of semiconducting and metallic systems. K-ART is parallelized in a non-trivial way: A master process uses several worker processes to perform independent event searches for possible events, while all bookkeeping and the actual simulation is performed by the master process. Depending on the complexity of the system studied, the parallelization scales well for tens to more than one hundred processes. For dealing with large systems, we present a near order 1 implementation. Techniques such as Verlet lists, cell decomposition and partial force calculations are implemented, and the CPU time per time step scales sublinearly with the number of particles, providing an efficient use of computational resources.

AB - We present two major optimizations for the kinetic Activation-Relaxation Technique (k-ART), an off-lattice self-learning kinetic Monte Carlo (KMC) algorithm with on-the-fly event search THAT has been successfully applied to study a number of semiconducting and metallic systems. K-ART is parallelized in a non-trivial way: A master process uses several worker processes to perform independent event searches for possible events, while all bookkeeping and the actual simulation is performed by the master process. Depending on the complexity of the system studied, the parallelization scales well for tens to more than one hundred processes. For dealing with large systems, we present a near order 1 implementation. Techniques such as Verlet lists, cell decomposition and partial force calculations are implemented, and the CPU time per time step scales sublinearly with the number of particles, providing an efficient use of computational resources.

UR - http://www.scopus.com/inward/record.url?scp=84858395616&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84858395616&partnerID=8YFLogxK

U2 - 10.1088/1742-6596/341/1/012007

DO - 10.1088/1742-6596/341/1/012007

M3 - Article

VL - 341

JO - Journal of Physics: Conference Series

JF - Journal of Physics: Conference Series

SN - 1742-6588

IS - 1

M1 - 012007

ER -