国产超级计算机实现10万核FDTD并行计算

doi:10.3969/j.issn.1001-2400.2015.05.015

Abstract

Abstract:

The massively parallel Finite-Difference Time-Domain(FDTD) computation using 100000 CPU cores is firstly implemented. Test results show that the parallel efficiency can reach up to 65% on 10240 CPU cores with 128 CPU cores as the benchmark. The research results in this paper indicate that the complicated electromagnetic problems can be solved accurately and efficiently using the method on the home-made supercomputer platform.

Key words: parallel algorithms, 100000 CPU cores, finite-difference-time-domain, computational electromagnetics

JIANG Shugang;LIN Zhongchao;ZHANG Yu;WEI Bing;ZHAO Xunwang. Parallel FDTD computation of hundred thousand cores on a home-made supercomputer[J].J4, 2015, 42(5): 86-91.

References

［1］ Yee K S. Numerical Solution of Initial Boundary Value Problems Involving Maxwell Equations in Isotropic Media ［J］. IEEE Transactions on Antennas and Propagation, 1966, 14(3): 302-307.
［2］葛德彪, 闫玉波. 电磁场时域有限差分方法［M］. 3版. 西安: 西安电子科技大学出版社, 2011.
［3］ Taflove A. Computational Electrodynamics: the Finite-difference Time-domain Method ［M］. Norwood: Artech House, 2000.
［4］张玉. 电磁场并行计算［M］. 西安: 西安电子科技大学出版社, 2006.
［5］雷继兆, 梁昌洪, 张玉. 并行FDTD结合服务器分析电大电磁问题［J］. 西安电子科技大学学报, 2009, 36(5): 846-850.
Lei Jizhao, Liang Changhong, Zhang Yu. Solving Electrically Large EM Problems Using Parallel FDTD and HP Blade Server ［J］. Journal of Xidian University, 2009, 36(5): 846-850.
［6］ He Z L, Huang K, Zhang Y, et al. Study on High Performance of MPI-based Parallel FDTD from Work-station to Super Computer Platform ［J］. International Journal of Antennas and Propagation, 2012, 2012: 1-7.
［7］ Jiang S G, Lü Z F, Zhang Y, et al. Analysis of Parallel Performance of MPI Based Parallel FDTD on Supercomputer ［C］//IET Conference Publications. Stevenage: Institution of Engineering and Technology, 2013: 1-4.
［8］ Guo X M, Guo Q X, Zhao W, et al. Parallel FDTD Simulation Using NUMA Acceleration Technique ［J］. Progress in Electromagnetics Research Letters, 2012, 28: 1-8.
［9］徐磊, 徐莹, 蒋荣琳, 等. GPU集群上的三维UPML-FDTD算法的实现及优化［J］. 计算机工程与科学, 2013, 2013(11): 160-167.
Xu Lei, Xu Ying, Jiang Ronglin, et al. Implementation and Optimization of Three-dimensional UPML-FDTD Algorithm on GPU Cluster ［J］. Computer Engineering and Science, 2013, 2013(11): 160-167.
［10］张立红, 余文华, 杨小玲. 加速并行时域有限差分仿真的新方法［J］. 电波科学学报, 2012, 27(1): 56-60.
Zhang Lihong, Yu Wenhua, Yang Xiaoling. New Acceleration Technique for Parallel FDTD Simulation ［J］. Chinese Journal of Radio Science, 2012, 27(1): 56-60.
［11］ Wang J, Yin W Y, Xia Y S. A Novel Conformal Surface Current Technique for Large Problems Based on High-performance Parallel FDTD Method ［J］. IEEE Antennas and Wireless Propagation Letters, 2013, 12: 11-14.
［12］ Hemmi T, Costen F, Garcia S, et al. Efficient Parallel LOD-FDTD Method for Debye-dispersive Media ［J］. IEEE Transactions on Antennas and Propagation, 2014, 62(3): 1330-1338.
［13］ Yu W H, Yang X L, Liu Y J, et al. A New Direction in Computational Electromagnetics: Solving Large Problems Using the Parallel FDTD on the BlueGene/L Supercomputer Providing Teraflop-level Performance ［J］. IEEE Antennas and Propagation Magazine, 2008, 50(2): 26-44.
［14］江树刚, 张玉, 赵勋旺, 等. 并行FDTD分析机载超短波天线［C］//2013 年全国微波毫米波会议论文集. 北京: 电子工业出版社, 2013: 211-214.
［15］ Jiang S G, Wei B, Zhang Y. Analysis of Near-field Characteristic of Airborne Antenna ［C］//Cross Strait Quad-regional Radio Science and Wireless Technology Conference. Washington: IEEE Computer Society, 2013: 160-162.

[1]	SU Qin, GU Zongjing, LIN Zhongchao, ZHAO Xunwang, ZHANG Yu. Analysis of scattering characteristics of the one-thousand-wavelength object using the parallel IE-NDDM [J]. Journal of Xidian University, 2018, 45(6): 7-13.
[2]	JIANG Xuesong;JIANG Shugang;ZHANG Yu;ZHAO Xunwang. Parallel FDTD computation of million cores with the domestically-made many-core supercomputer [J]. Journal of Xidian University, 2017, 44(6): 65-69+128.
[3]	LIN Zhongchao;CHEN Yan;ZHANG Yu;JIANG Shugang;ZHAO Xunwang;L Zhaofeng. Study of the parallel higher-order MoM on a domestically-made CPU platform [J]. J4, 2015, 42(3): 43-47.
[4]	WU Huixin;WANG Bing;CHAI Zhengyi. Spectrum allocation algorithm based on parallel immune optimization [J]. J4, 2014, 41(4): 111-115+136.
[5]	LEI Ji-zhao;LIANG Chang-hong;ZHANG Yu. Solving electrically large EM problems using parallel FDTD and HP blade server [J]. J4, 2009, 36(5): 846-850.
[6]	ZHU Hu-ming;JIAO Li-cheng. Parallel immune clonal selection for feature selection [J]. J4, 2008, 35(5): 853-857.

Parallel FDTD computation of hundred thousand cores on a home-made supercomputer

PDF (PC)

Like

Knowledge

Cited

Abstract

Cite this article

share this article

References

Related Articles 6

Metrics

Comments

Recommended 6