Multiple-GPU parallelization of three-dimensional material point method based on single-root complex

doi:10.1002/nme.6906

	Multiple-GPU parallelization of three-dimensional material point method based on single-root complex
	Dong, Youkou 3,4; Cui, Lan 2; Zhang, Xue 1,4
刊名	INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN ENGINEERING
	2022-01-04
页码	24
关键词	cone penetration test material point method mini slump test parallel computation
ISSN号	0029-5981
DOI	10.1002/nme.6906
英文摘要	As one of the arbitrary Lagrangian-Eulerian methods, the material point method (MPM) owns intrinsic advantages in simulation of large deformation problems by combining the merits of the Lagrangian and Eulerian approaches. Significant computational intensity is involved in the calculations of the MPM due to its very fine mesh needed to achieve a sufficiently high accuracy. A new multiple-GPU parallel strategy is developed based on a single-root complex architecture of the computer purely within a CUDA environment. Peer-to-Peer (P2P) communication between the GPUs is performed to exchange the information of the crossing particles and ghost element nodes, which is faster than the heavy send/receive operations between different computers through the infiniBand network. Domain decomposition is performed to split the whole computational task over the GPUs with a number of subdomains. The computations within each subdomain are allocated on a corresponding GPU using an enhanced "Particle-List" scheme to tackle the data race during the interpolation from associated particles to common nodes. The acceleration effect of the parallelization is evaluated with two benchmarks cases, mini-slump test after a dam break and cone penetration test in clay, where the maximum speedups with 1 and 8 GPUs are 88 and 604, respectively.
资助项目	National Natural Science Foundations of China[51909248] ; Open Research Fund of State Key Laboratory of Coastal and Offshore Engineering, Dalian University of Technology[LP2012] ; NVIDIA Corporation
WOS研究方向	Engineering ; Mathematics
语种	英语
出版者	WILEY
WOS记录号	WOS:000738369800001
内容类型	期刊论文
源URL	[http://119.78.100.198/handle/2S6PX9GI/30815]
专题	中科院武汉岩土力学所
通讯作者	Cui, Lan
作者单位	1.Univ Liverpool, Dept Civil Engn & Ind Design, Liverpool, Merseyside, England 2.Chinese Acad Sci, Inst Rock & Soil Mech, State Key Lab Geomech & Geotech Engn, Wuhan 430071, Peoples R China 3.Dalian Univ Technol, State Key Lab Coastal & Offshore Engn, Dalian, Peoples R China 4.China Univ Geosci, Coll Marine Sci & Technol, Wuhan, Peoples R China
推荐引用方式 GB/T 7714	Dong, Youkou,Cui, Lan,Zhang, Xue. Multiple-GPU parallelization of three-dimensional material point method based on single-root complex[J]. INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN ENGINEERING,2022:24.
APA	Dong, Youkou,Cui, Lan,&Zhang, Xue.(2022).Multiple-GPU parallelization of three-dimensional material point method based on single-root complex.INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN ENGINEERING,24.
MLA	Dong, Youkou,et al."Multiple-GPU parallelization of three-dimensional material point method based on single-root complex".INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN ENGINEERING (2022):24.