Implementation of a lattice Boltzmann method for large eddy simulation on multiple GPUs

Qinjian Li; Chengwen Zhong; Kai Li; Guangyong Zhang; Xiaowei Lu; Qing Zhang; Kaiyong Zhao; Xiaowen CHU

doi:10.1109/HPCC.2012.115

Implementation of a lattice Boltzmann method for large eddy simulation on multiple GPUs

Qinjian Li^*, Chengwen Zhong, Kai Li, Guangyong Zhang, Xiaowei Lu, Qing Zhang, Kaiyong Zhao, Xiaowen CHU

^*Corresponding author for this work

Department of Computer Science

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

3 Citations (Scopus)

Abstract

Recently, the Graphic Processor Unit (GPU) has evolved into a highly parallel, multithreaded, many-core processor with tremendous computational horsepower and very high memory bandwidth. To improve the simulation efficiency of complex flow phenomena in the field of computational fluid dynamics, a CUDA-based simulation algorithm of large eddy simulation using multiple GPUs is proposed. Our implementation adopted the "collision after propagation" scheme and performed the propagation process by global memory reading transactions. The working set is split up into equal sub-domains and assigned to each GPU for simplicity. Using recently released hardware, up to four GPUs can be controlled by a single CPU thread and run in parallel. The results show that our multi-GPU implementation could perform simulations on a rather large scale (meshes: 10240x10240) even using double-precision floating point calculation and achieved 190X speedup over the sequential implementation on CPU.

Original language	English
Title of host publication	Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012
Pages	818-823
Number of pages	6
DOIs	https://doi.org/10.1109/HPCC.2012.115
Publication status	Published - 2012
Event	14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012 - Liverpool, United Kingdom Duration: 25 Jun 2012 → 27 Jun 2012 https://ieeexplore.ieee.org/xpl/conhome/6331801/proceeding

Publication series

Name	Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012

Conference

Conference	14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012
Country/Territory	United Kingdom
City	Liverpool
Period	25/06/12 → 27/06/12
Internet address	https://ieeexplore.ieee.org/xpl/conhome/6331801/proceeding

Scopus Subject Areas

Software

User-Defined Keywords

Large eddy simulation
Lattice Boltzmann method
Multi-GPU Computing
Parellel computing

Access to Document

10.1109/HPCC.2012.115

Cite this

Li, Q., Zhong, C., Li, K., Zhang, G., Lu, X., Zhang, Q., Zhao, K., & CHU, X. (2012). Implementation of a lattice Boltzmann method for large eddy simulation on multiple GPUs. In Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012 (pp. 818-823). Article 6332253 (Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012). https://doi.org/10.1109/HPCC.2012.115

Li, Qinjian ; Zhong, Chengwen ; Li, Kai et al. / Implementation of a lattice Boltzmann method for large eddy simulation on multiple GPUs. Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012. 2012. pp. 818-823 (Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012).

@inproceedings{7d8881940dff4be79ff8091f49ee162d,

title = "Implementation of a lattice Boltzmann method for large eddy simulation on multiple GPUs",

abstract = "Recently, the Graphic Processor Unit (GPU) has evolved into a highly parallel, multithreaded, many-core processor with tremendous computational horsepower and very high memory bandwidth. To improve the simulation efficiency of complex flow phenomena in the field of computational fluid dynamics, a CUDA-based simulation algorithm of large eddy simulation using multiple GPUs is proposed. Our implementation adopted the {"}collision after propagation{"} scheme and performed the propagation process by global memory reading transactions. The working set is split up into equal sub-domains and assigned to each GPU for simplicity. Using recently released hardware, up to four GPUs can be controlled by a single CPU thread and run in parallel. The results show that our multi-GPU implementation could perform simulations on a rather large scale (meshes: 10240x10240) even using double-precision floating point calculation and achieved 190X speedup over the sequential implementation on CPU.",

keywords = "Large eddy simulation, Lattice Boltzmann method, Multi-GPU Computing, Parellel computing",

author = "Qinjian Li and Chengwen Zhong and Kai Li and Guangyong Zhang and Xiaowei Lu and Qing Zhang and Kaiyong Zhao and Xiaowen CHU",

note = "Copyright: Copyright 2012 Elsevier B.V., All rights reserved.; 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012 ; Conference date: 25-06-2012 Through 27-06-2012",

year = "2012",

doi = "10.1109/HPCC.2012.115",

language = "English",

isbn = "9780769547497",

series = "Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012",

pages = "818--823",

booktitle = "Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012",

url = "https://ieeexplore.ieee.org/xpl/conhome/6331801/proceeding",

}

Li, Q, Zhong, C, Li, K, Zhang, G, Lu, X, Zhang, Q, Zhao, K & CHU, X 2012, Implementation of a lattice Boltzmann method for large eddy simulation on multiple GPUs. in Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012., 6332253, Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012, pp. 818-823, 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012, Liverpool, United Kingdom, 25/06/12. https://doi.org/10.1109/HPCC.2012.115

Implementation of a lattice Boltzmann method for large eddy simulation on multiple GPUs. / Li, Qinjian; Zhong, Chengwen; Li, Kai et al.
Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012. 2012. p. 818-823 6332253 (Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012).

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

TY - GEN

T1 - Implementation of a lattice Boltzmann method for large eddy simulation on multiple GPUs

AU - Li, Qinjian

AU - Zhong, Chengwen

AU - Li, Kai

AU - Zhang, Guangyong

AU - Lu, Xiaowei

AU - Zhang, Qing

AU - Zhao, Kaiyong

AU - CHU, Xiaowen

PY - 2012

Y1 - 2012

N2 - Recently, the Graphic Processor Unit (GPU) has evolved into a highly parallel, multithreaded, many-core processor with tremendous computational horsepower and very high memory bandwidth. To improve the simulation efficiency of complex flow phenomena in the field of computational fluid dynamics, a CUDA-based simulation algorithm of large eddy simulation using multiple GPUs is proposed. Our implementation adopted the "collision after propagation" scheme and performed the propagation process by global memory reading transactions. The working set is split up into equal sub-domains and assigned to each GPU for simplicity. Using recently released hardware, up to four GPUs can be controlled by a single CPU thread and run in parallel. The results show that our multi-GPU implementation could perform simulations on a rather large scale (meshes: 10240x10240) even using double-precision floating point calculation and achieved 190X speedup over the sequential implementation on CPU.

AB - Recently, the Graphic Processor Unit (GPU) has evolved into a highly parallel, multithreaded, many-core processor with tremendous computational horsepower and very high memory bandwidth. To improve the simulation efficiency of complex flow phenomena in the field of computational fluid dynamics, a CUDA-based simulation algorithm of large eddy simulation using multiple GPUs is proposed. Our implementation adopted the "collision after propagation" scheme and performed the propagation process by global memory reading transactions. The working set is split up into equal sub-domains and assigned to each GPU for simplicity. Using recently released hardware, up to four GPUs can be controlled by a single CPU thread and run in parallel. The results show that our multi-GPU implementation could perform simulations on a rather large scale (meshes: 10240x10240) even using double-precision floating point calculation and achieved 190X speedup over the sequential implementation on CPU.

KW - Large eddy simulation

KW - Lattice Boltzmann method

KW - Multi-GPU Computing

KW - Parellel computing

UR - http://www.scopus.com/inward/record.url?scp=84870422464&partnerID=8YFLogxK

U2 - 10.1109/HPCC.2012.115

DO - 10.1109/HPCC.2012.115

M3 - Conference proceeding

AN - SCOPUS:84870422464

SN - 9780769547497

T3 - Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012

SP - 818

EP - 823

BT - Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012

T2 - 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012

Y2 - 25 June 2012 through 27 June 2012

ER -

Li Q, Zhong C, Li K, Zhang G, Lu X, Zhang Q et al. Implementation of a lattice Boltzmann method for large eddy simulation on multiple GPUs. In Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012. 2012. p. 818-823. 6332253. (Proceedings of the 14th IEEE International Conference on High Performance Computing and Communications, HPCC-2012 - 9th IEEE International Conference on Embedded Software and Systems, ICESS-2012). doi: 10.1109/HPCC.2012.115

Implementation of a lattice Boltzmann method for large eddy simulation on multiple GPUs

Abstract

Publication series

Conference

Scopus Subject Areas

User-Defined Keywords

Access to Document

Other files and links

Fingerprint

Cite this