Application of an Improved GPU Acceleration Strategy for the Smoothed Particle Hydrodynamics Method

doi:10.16183/j.cnki.jsjtu.2022.209

Abstract

Abstract:

In order to solve the problem of graphics processing unit (GPU) memory access conflicts possibly caused by the disorder of particles and enhance the computation efficiency, an improved GPU acceleration strategy is proposed by establishing particle reorder technology. The acceleration strategy is applied to the smoothed particle hydrodynamics (SPH) method to simulate the dam breaking with obstacles in three dimensions, and the algorithm is verified by comparing with the experimental results, which obtained a high calculation accuracy. Based on this benchmark example of the SPH, the studies on the effect of particle renumbering and the solution efficiency of the algorithm are conducted by comparing the simulations of different hardware facilities. The results indicate that the particle reorder technology can ensure a stable single-step running time, and can effectively solve the problem of graphic card memory access conflicts that commonly exist in the GPU-SPH algorithm. Furthermore, the GPU parallel algorithm can greatly improve the solution efficiency of the SPH method, and with the increase of particle number, the advantage of drastically reducing the computation time becomes more obvious. The method proposed in this paper provides the possibility to expand the application of the SPH method to solve 3D numerical simulations.

Key words: smoothed particle hydrodynamics (SPH), parallel computing, dam breaking, computational efficiency

CLC Number:

U661.1

GUAN Yanmin, YANG Caihong, KANG Zhuang, ZHOU Li. Application of an Improved GPU Acceleration Strategy for the Smoothed Particle Hydrodynamics Method[J]. Journal of Shanghai Jiao Tong University, 2023, 57(8): 981-987.

Figures/Tables 10

Fig.1

Fig.2

Fig.3

Fig.4

Fig.5

Tab.1

Fig.6

Fig.7

Fig.8

Tab.2

References 16

[1]	刘肃肃, 胡祎乐, 余音. 基于GPU 的近场动力学模拟的并行化方法[J]. 上海交通大学学报, 2016, 50(9): 1362-1367.
	LIU Susu, HU Yile, YU Yin. Parallel computing method of peridynamic models based on GPU[J]. Journal of Shanghai Jiao Tong University, 2016, 50(9): 1362-1367.
[2]	HARADA T, KOSHIZUKA S, KAWAGUCHI Y. Smoothed particle hydrodynamics on GPUs[C]//Computer Graphics International. Petropolis, Brazil: Computer Graphics Society, 2007, 40: 63-70.
[3]	CRESPO A C, DOMINGUEZ J M, BARREIRO A, et al. GPUs, a new tool of acceleration in CFD: Efficiency and reliability on smoothed particle hydrodynamics methods[J]. PloS One, 2011, 6(6): 1-13.
[4]	HE Y, BAYLY A E, HASSANPOUR A, et al. A GPU-based coupled SPH-DEM method for particle-fluid flow with free surfaces[J]. Powder Technology, 2018, 338: 548-562. doi: 10.1016/j.powtec.2018.07.043 URL
[5]	XIA X, LIANG Q. A GPU-accelerated smoothed particle hydrodynamics (SPH) model for the shallow water equations[J]. Environmental Modelling & Software, 2016, 75: 28-43.
[6]	XIONG Q, LI B, XU J. GPU-accelerated adaptive particle splitting and merging in SPH[J]. Computer Physics Communication, 2013, 184 (7): 1701-1707. doi: 10.1016/j.cpc.2013.02.021 URL
[7]	徐锋. 基于众核架构的并行SPH 算法的研究与实现[D]. 上海: 上海交通大学, 2013.
	XU Feng. Research and implementation of the smoothed particle hydrodynamics algorithm based on multi-core architecture[D]. Shanghai: Shanghai Jiao Tong University, 2013.
[8]	金善勤, 郑兴, 段文洋. 基于GPU 并行的改进SPH 方法对黏性流场的模拟[J]. 哈尔滨工程大学学报, 2015, 36(8): 1011-1018.
	JIN Shanqin, ZHENG Xing, DUAN Wenyang. Viscosity flow simulation using improved SPH method based on GPU parallel calculation[J]. Journal of Harbin Engineering University, 2015, 36(8): 1011-1018.
[9]	杨志国, 黄兴, 郑兴, 等. GPU在SPH方法模拟溃坝问题的应用研究[J]. 哈尔滨工程大学学报, 2014, 35(6): 661-666.
	YANG Zhiguo, HUANG Xing, ZHENG Xing, et al. The application research of GPU in the SPH method to simulate the dam breaking problem[J]. Journal of Harbin Engineering University, 2014, 35(6): 661-666.
[10]	车庆首, 李传文, 张轶, 等. GAPI: GPU 加速的移动对象并行索引方法[J]. 计算机科学与探索, 2017, 11(11): 1713-1722. doi: 10.3778/j.issn.1673-9418.1608038
	CHE Qingshou, LI Chuanwen, ZHANG Yi, et al. GAPI: GPU accelerated parallel method for indexing moving objects[J]. Journal of Frontiers of Computer Science and Technology, 2017, 11(11): 1713-1722.
[11]	IHMSEN M, AKINCI N, BECKER M, et al. A parallel SPH implementation on multi-core CPUs[J]. Computer Graphics Forum, 2011, 30: 99-112. doi: 10.1111/j.1467-8659.2010.01832.x URL
[12]	聂霄. 不可压缩SPH流体的真实感模拟及其加速技术研究[D]. 成都: 电子科技大学, 2015.
	NIE Xiao. Study on realistic simulation and acceleration techniques of incompressible SPH fluids[D]. Chengdu: University of Electronic Science and Technology of China, 2015.
[13]	WENDLAND H. Piecewise polynomial, positive definite and compactly supported radial functions of minimal degree[J]. Advances in computational Mathematics, 1995, 4(1): 389-396. doi: 10.1007/BF02123482 URL
[14]	MONAGHAN J J, GINGOLD R A. Shock simulation by the particle method SPH[J]. Journal of Computational Physics, 1983, 52(2): 374-389. doi: 10.1016/0021-9991(83)90036-0 URL
[15]	MONAGHAN J J. Particle methods for hydrodynamics[J]. Computer Physics Report, 1985, 3(2): 71-124. doi: 10.1016/0167-7977(85)90010-3 URL
[16]	KLEEFSMAN K M T, FEKKEN G, VELDMAN A E P, et al. A volume-of-fluid based simulation method for wave impact problems[J]. Journal of Computational Physics, 2005, 206: 363-393. doi: 10.1016/j.jcp.2004.12.007 URL

设备类型	型号	核数	主频/GHz
CPU	Intel Core i9-10900F	10	2.8
	Xeon?Gold 6268CL	48	2.8
GPU	GTX 1660 super	1 408	1.785
	GeForce RTX 2080	2 944	1. 4
	GeForce RTX 3080Ti	10 240	1.67

硬件	核数比	效率比	核数/效率
Intel Core i9-10900F	1	1	1
Xeon?Gold 6268CL	4.8	1.73	2.78
GTX 1660 super	140.8	31.05	4.53
GeForce RTX 2080	294.4	59.04	4.99
GeForce RTX 3080Ti	1 024	104.8	9.76