Simulating spin models on GPU

Martin Weigel

doi:10.1016/j.cpc.2010.10.031

Simulating spin models on GPU

Martin Weigel

Research output: Contribution to journal › Article

43 Citations (Scopus)

Abstract

Over the last couple of years it has been realized that the vast computational power of graphics processing units (GPUs) could be harvested for purposes other than the video game industry. This power, which at least nominally exceeds that of current CPUs by large factors, results from the relative simplicity of the GPU architectures as compared to CPUs, combined with a large number of parallel processing units on a single chip. To benefit from this setup for general computing purposes, the problems at hand need to be prepared in a way to profit from the inherent parallelism and hierarchical structure of memory accesses. In this contribution I discuss the performance potential for simulating spin models, such as the Ising model, on GPU as compared to conventional simulations on CPU.

Original language	English
Pages (from-to)	1833–1836
Journal	Computer Physics Communications
Volume	182
Issue number	9
DOIs	https://doi.org/10.1016/j.cpc.2010.10.031
Publication status	Published - 2011

Bibliographical note

The full text is available free from the link given. The published version can be found at http://dx.doi.org10.1016/j.cpc.2010.10.031. Please note the author was working at the Institut für Physik, Johannes Gutenberg-Universität Mainz at the time of publication.

Keywords

Monte Carlo simulations
GPU
spin models

Access to Document

10.1016/j.cpc.2010.10.031

http://arxiv.org/abs/1006.3865

Cite this

@article{523b63fd614d4ff6a6201f5ea77f59b4,

title = "Simulating spin models on GPU",

abstract = "Over the last couple of years it has been realized that the vast computational power of graphics processing units (GPUs) could be harvested for purposes other than the video game industry. This power, which at least nominally exceeds that of current CPUs by large factors, results from the relative simplicity of the GPU architectures as compared to CPUs, combined with a large number of parallel processing units on a single chip. To benefit from this setup for general computing purposes, the problems at hand need to be prepared in a way to profit from the inherent parallelism and hierarchical structure of memory accesses. In this contribution I discuss the performance potential for simulating spin models, such as the Ising model, on GPU as compared to conventional simulations on CPU.",

keywords = "Monte Carlo simulations, GPU, spin models",

author = "Martin Weigel",

note = "The full text is available free from the link given. The published version can be found at http://dx.doi.org10.1016/j.cpc.2010.10.031. Please note the author was working at the Institut f{\"u}r Physik, Johannes Gutenberg-Universit{\"a}t Mainz at the time of publication.",

year = "2011",

doi = "10.1016/j.cpc.2010.10.031",

language = "English",

volume = "182",

pages = "1833–1836",

journal = "Computer Physics Communications",

issn = "0010-4655",

publisher = "Elsevier",

number = "9",

}

TY - JOUR

T1 - Simulating spin models on GPU

AU - Weigel, Martin

N1 - The full text is available free from the link given. The published version can be found at http://dx.doi.org10.1016/j.cpc.2010.10.031. Please note the author was working at the Institut für Physik, Johannes Gutenberg-Universität Mainz at the time of publication.

PY - 2011

Y1 - 2011

N2 - Over the last couple of years it has been realized that the vast computational power of graphics processing units (GPUs) could be harvested for purposes other than the video game industry. This power, which at least nominally exceeds that of current CPUs by large factors, results from the relative simplicity of the GPU architectures as compared to CPUs, combined with a large number of parallel processing units on a single chip. To benefit from this setup for general computing purposes, the problems at hand need to be prepared in a way to profit from the inherent parallelism and hierarchical structure of memory accesses. In this contribution I discuss the performance potential for simulating spin models, such as the Ising model, on GPU as compared to conventional simulations on CPU.

AB - Over the last couple of years it has been realized that the vast computational power of graphics processing units (GPUs) could be harvested for purposes other than the video game industry. This power, which at least nominally exceeds that of current CPUs by large factors, results from the relative simplicity of the GPU architectures as compared to CPUs, combined with a large number of parallel processing units on a single chip. To benefit from this setup for general computing purposes, the problems at hand need to be prepared in a way to profit from the inherent parallelism and hierarchical structure of memory accesses. In this contribution I discuss the performance potential for simulating spin models, such as the Ising model, on GPU as compared to conventional simulations on CPU.

KW - Monte Carlo simulations

KW - GPU

KW - spin models

U2 - 10.1016/j.cpc.2010.10.031

DO - 10.1016/j.cpc.2010.10.031

M3 - Article

VL - 182

SP - 1833

EP - 1836

JO - Computer Physics Communications

JF - Computer Physics Communications

SN - 0010-4655

IS - 9

ER -

Simulating spin models on GPU

Abstract

Bibliographical note

Keywords

Access to Document

Fingerprint

Cite this