GPU-accelerated phase field simulation of directional solidification

Ang Gao; Yansu Hu; Zhijun Wang; Dejun Mu; Junjie Li; Jincheng Wang

doi:10.1007/s11431-014-5541-1

GPU-accelerated phase field simulation of directional solidification

Ang Gao, Yansu Hu, Zhijun Wang, Dejun Mu, Junjie Li, Jincheng Wang

School of Materials Sience and Engineering

Northwestern Polytechnical University Xian

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

5 Scopus citations

Abstract

The phase field simulation has been actively studied as a powerful method to investigate the microstructural evolution during the solidification. However, it is a great challenge to perform the phase field simulation in large length and time scale. The developed graphics processing unit (GPU) calculation is used in the phase filed simulation, greatly accelerating the calculation efficiency. The results show that the computation with GPU is about 36 times faster than that with a single Central Processing Unit (CPU) core. It provides the feasibility of the GPU-accelerated phase field simulation on a desktop computer. The GPU-accelerated strategy will bring a new opportunity to the application of phase field simulation.

Original language	English
Pages (from-to)	1191-1197
Number of pages	7
Journal	Science China Technological Sciences
Volume	57
Issue number	6
DOIs	https://doi.org/10.1007/s11431-014-5541-1
State	Published - Jun 2014

Keywords

computer unified device architecture (CUDA)
directional solidification
graphics processing unit (GPU) acceleration
phase field simulation
speed-up ratio

Access to Document

10.1007/s11431-014-5541-1

Cite this

@article{b744ba70be8c4e87ae133281ef973cd3,

title = "GPU-accelerated phase field simulation of directional solidification",

abstract = "The phase field simulation has been actively studied as a powerful method to investigate the microstructural evolution during the solidification. However, it is a great challenge to perform the phase field simulation in large length and time scale. The developed graphics processing unit (GPU) calculation is used in the phase filed simulation, greatly accelerating the calculation efficiency. The results show that the computation with GPU is about 36 times faster than that with a single Central Processing Unit (CPU) core. It provides the feasibility of the GPU-accelerated phase field simulation on a desktop computer. The GPU-accelerated strategy will bring a new opportunity to the application of phase field simulation.",

keywords = "computer unified device architecture (CUDA), directional solidification, graphics processing unit (GPU) acceleration, phase field simulation, speed-up ratio",

author = "Ang Gao and Yansu Hu and Zhijun Wang and Dejun Mu and Junjie Li and Jincheng Wang",

year = "2014",

month = jun,

doi = "10.1007/s11431-014-5541-1",

language = "英语",

volume = "57",

pages = "1191--1197",

journal = "Science China Technological Sciences",

issn = "1674-7321",

publisher = "Science China Press ",

number = "6",

}

TY - JOUR

T1 - GPU-accelerated phase field simulation of directional solidification

AU - Gao, Ang

AU - Hu, Yansu

AU - Wang, Zhijun

AU - Mu, Dejun

AU - Li, Junjie

AU - Wang, Jincheng

PY - 2014/6

Y1 - 2014/6

N2 - The phase field simulation has been actively studied as a powerful method to investigate the microstructural evolution during the solidification. However, it is a great challenge to perform the phase field simulation in large length and time scale. The developed graphics processing unit (GPU) calculation is used in the phase filed simulation, greatly accelerating the calculation efficiency. The results show that the computation with GPU is about 36 times faster than that with a single Central Processing Unit (CPU) core. It provides the feasibility of the GPU-accelerated phase field simulation on a desktop computer. The GPU-accelerated strategy will bring a new opportunity to the application of phase field simulation.

AB - The phase field simulation has been actively studied as a powerful method to investigate the microstructural evolution during the solidification. However, it is a great challenge to perform the phase field simulation in large length and time scale. The developed graphics processing unit (GPU) calculation is used in the phase filed simulation, greatly accelerating the calculation efficiency. The results show that the computation with GPU is about 36 times faster than that with a single Central Processing Unit (CPU) core. It provides the feasibility of the GPU-accelerated phase field simulation on a desktop computer. The GPU-accelerated strategy will bring a new opportunity to the application of phase field simulation.

KW - computer unified device architecture (CUDA)

KW - directional solidification

KW - graphics processing unit (GPU) acceleration

KW - phase field simulation

KW - speed-up ratio

UR - http://www.scopus.com/inward/record.url?scp=84903648161&partnerID=8YFLogxK

U2 - 10.1007/s11431-014-5541-1

DO - 10.1007/s11431-014-5541-1

M3 - 文章

AN - SCOPUS:84903648161

SN - 1674-7321

VL - 57

SP - 1191

EP - 1197

JO - Science China Technological Sciences

JF - Science China Technological Sciences

IS - 6

ER -

GPU-accelerated phase field simulation of directional solidification

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this