Massively parallel Wang-Landau sampling on multiple GPUs

Junqi Yin; D. P. Landau

doi:10.1016/j.cpc.2012.02.023

Massively parallel Wang-Landau sampling on multiple GPUs

Junqi Yin, D. P. Landau

Analytics and AI Methods at Scale

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

30 Scopus citations

Abstract

Wang-Landau sampling is implemented on the Graphics Processing Unit (GPU) with the Compute Unified Device Architecture (CUDA). Performances on three different GPU cards, including the new generation Fermi architecture card, are compared with that on a Central Processing Unit (CPU). The parameters for massively parallel Wang-Landau sampling are tuned in order to achieve fast convergence. For simulations of the water cluster systems, we obtain an average of over 50 times speedup for a given workload.

Original language	English
Pages (from-to)	1568-1573
Number of pages	6
Journal	Computer Physics Communications
Volume	183
Issue number	8
DOIs	https://doi.org/10.1016/j.cpc.2012.02.023
State	Published - Aug 2012

Keywords

GPU computing
Phase transition
Wang-Landau sampling
Water clusters

Access to Document

10.1016/j.cpc.2012.02.023

Cite this

@article{d410bfb3a64547cfb9b8a69a14c69ccc,

title = "Massively parallel Wang-Landau sampling on multiple GPUs",

abstract = "Wang-Landau sampling is implemented on the Graphics Processing Unit (GPU) with the Compute Unified Device Architecture (CUDA). Performances on three different GPU cards, including the new generation Fermi architecture card, are compared with that on a Central Processing Unit (CPU). The parameters for massively parallel Wang-Landau sampling are tuned in order to achieve fast convergence. For simulations of the water cluster systems, we obtain an average of over 50 times speedup for a given workload.",

keywords = "GPU computing, Phase transition, Wang-Landau sampling, Water clusters",

author = "Junqi Yin and Landau, {D. P.}",

year = "2012",

month = aug,

doi = "10.1016/j.cpc.2012.02.023",

language = "English",

volume = "183",

pages = "1568--1573",

journal = "Computer Physics Communications",

issn = "0010-4655",

publisher = "Elsevier",

number = "8",

}

TY - JOUR

T1 - Massively parallel Wang-Landau sampling on multiple GPUs

AU - Yin, Junqi

AU - Landau, D. P.

PY - 2012/8

Y1 - 2012/8

N2 - Wang-Landau sampling is implemented on the Graphics Processing Unit (GPU) with the Compute Unified Device Architecture (CUDA). Performances on three different GPU cards, including the new generation Fermi architecture card, are compared with that on a Central Processing Unit (CPU). The parameters for massively parallel Wang-Landau sampling are tuned in order to achieve fast convergence. For simulations of the water cluster systems, we obtain an average of over 50 times speedup for a given workload.

AB - Wang-Landau sampling is implemented on the Graphics Processing Unit (GPU) with the Compute Unified Device Architecture (CUDA). Performances on three different GPU cards, including the new generation Fermi architecture card, are compared with that on a Central Processing Unit (CPU). The parameters for massively parallel Wang-Landau sampling are tuned in order to achieve fast convergence. For simulations of the water cluster systems, we obtain an average of over 50 times speedup for a given workload.

KW - GPU computing

KW - Phase transition

KW - Wang-Landau sampling

KW - Water clusters

UR - http://www.scopus.com/inward/record.url?scp=84860304107&partnerID=8YFLogxK

U2 - 10.1016/j.cpc.2012.02.023

DO - 10.1016/j.cpc.2012.02.023

M3 - Article

AN - SCOPUS:84860304107

SN - 0010-4655

VL - 183

SP - 1568

EP - 1573

JO - Computer Physics Communications

JF - Computer Physics Communications

IS - 8

ER -

Massively parallel Wang-Landau sampling on multiple GPUs

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this