航天领域高性能并行计算研究进展

摘要/Abstract

摘要：

航天领域的大规模科学与工程问题的数值模拟既依赖于高性能并行计算的支撑，同时也是高性能并行计算发展的动力。综述了航天领域高性能并行计算的研究进展，对高性能并行计算环境进行简单介绍，对相关研究领域包括气动力、气动热、化学非平衡、结构强度、热防护、蒙特卡罗方法和湍流研究等进行分类和详细阐述；总结了航天领域高性能并行计算存在科学计算高并行效率和工程计算低实用价值、并行应用的多样性和缺少科学的并行方法的矛盾，并指出了进一步研究方向。

关键词: 高性能, 并行计算, 航天, 研究进展

Abstract:

The numerical simulation of the large scale science and engineering problem in the aerospace area relies on the support of HighPerformance Parallel Computing (HPPC) and is the engine of the development of HPPC. The paper gives a survey of recent progress of HPPC in the aerospace area and a brief introduction of the HPPC environment. The relative research areas of HPPC of aerospace, including aerodynamic, aerothermal, chemical nonequilibrium, structure, heat protection, Monte Carlo method and turbulence et al., are classified and described in detail. The dilemma of high parallel efficiency in scientific computing and low use in engineering computing, and the diversity of parallel applications and the lack of scientific parallel methods are concluded. Finally, the prospects of the future HPPC development in aerospace area are presented.

Key words: high performance;parallel computing;aerospace;recent progress

龚春叶1,2,3，包为民1,2，汤国建2，王玲1，孙学功1，刘杰3. 航天领域高性能并行计算研究进展[J]. J4, 2014, 36(09): 1629-1636.

GONG Chunye1,2,3，BAO Weimin1,2，TANG Guojian2，WANG Ling1，SUN Xuegong1，LIU Jie. Recent progress in high-performance
parallel computing of the aerospace area [J]. J4, 2014, 36(09): 1629-1636.

参考文献［40］

［1］	Ball D N. Contributions of CFD to the 787and Future Needs［EB/OL］.［20081020］.http://www.hpcuserforum.com/presentations/Tucson/Boeing%20Ball%20IDC%20pdf.pdf.
［2］	Asanovic K, Bodik R, Catanzaro B, et al. A view of the parallel computing landscape［J］. Communication of ACM. 2009, 52(10):5667.
［3］	Zhang Linbo, Chi Xuebi, Mo Zeyao, et al. Introduction of parallel computing［M］. Beijing:Tsinghua University Press, 2006.(in Chinese)
［4］	Meuer H,Strohmaier E,Dongarra J,et al.Home \| TOP500 supercomputing sites［EB/OL］.［20120620］.http://www.top500.org/.
［5］	ASCR D. International exascale software project roadmap 0.93［EB/OL］.［20100117］.http://www.exascale.org/.
［6］	Liu Chen. The numerical simulation method of complex combustion flow field［D］. Nanjing:Nanjing University of Aeronautics and Astronautics, 2009.(in Chinese)
［7］	Central goverment of China. The twelfth fiveyear plan for national economic and social development of the People’s Republic of China［EB/OL］.［20130301］.http://www.gov.cn/.(in Chinese)
［8］	National Nature Science Foundation of China(NSFC)“12th FiveYear” development rules of NSFC［EB/OL］.［20130301］.http://www.nsfc.gov.cn/.(in Chinese)
［9］	The Ministry of Science and Technology of China. Notice of the collection of highperformance computer and application services environments major project proposed by the 863 Plan in the 12th FiveYear Plan［EB/OL］.［20130301］.http://www.most.gov.cn/.(in Chinese)
［10］	Du Zhihui. Parallel programming techniques in highperformance computingMPI［M］. Beijing:Tsinghua University Press, 2001.(in Chinese)
［11］	NVIDIA Corporation. CUDA C Programming Guide Version 5.0［EB/OL］.［20130301］.http://www.nvidia.com/.
［12］	NASA.NASA advanced supercomputing division［EB/OL］.［20130301］.http://www.ssc.net.cn/.
［13］	Zhang Hanxin,Shen Mengyu.Computational fluid dynamics:Principles and applications of the differential method［M］ Beijing:National Defense Industry Press, 2003.(in Chinese)
［14］	Burstedde C, Ghattas O, Gurnis M, et al. Extremescale AMR［C］∥Proc of the 2010 ACM/IEEE International Conference for High Performance Computing, Networking, Storage and Analysis, 2010:112.
［15］	Hitchcock D. ASCR update:View from Washington［EB/OL］.［20130301］.http://www.science.energy.gov.
［16］	Sunderland A G,Ashworth M,Li N,et al. Towards petascale computing with parallel CFD codes［EB/OL］.［20130301］.http://www.hpcx.ac.uk.
［17］	Soga T, Musa A, Shimomura Y, et al. Performance evaluation of NEC SX9 using real science and engineering applications［C］∥Proc of the Conference on High Performance Computing Networking, Storage and Analysis,2009:1.
［18］	Sahni O, Zhou M, Shephard M, et al. Scalable implicit finite element solver for massively parallel processing with demonstration to 160k cores［C］∥Proc of the Conference on High Performance Computing Networking, Storage and Analysis, 2009:1.
［19］	Brandvik T, Pullan G. Acceleration of a 3D Euler solver using commodity graphics hardware［C］∥Proc of the 46th AIAA Aerospace Sciences Meeting and Exhibit, 2008:2008607.
［20］	Brandvik T, Pullan G. Acceleration of a twodimensional Euler flow solver using commodity graphics hardware［J］. Journal of Mechanical Engineering Science. 2007, 221(12):17451748.
［21］	Pan Sha. The research of the numrical simulation and massively parallel computing of hypersonic aerodynamic heating［D］. Changsha:National University of Defense Technology, 2010.(in Chinese)
［22］	Gourdain N, Montagnac M, Wlassow F, et al. Highperformance computing to simulate largescale industrial flows in multistage compressors［J］. International Journal of High Performance Computing Applications, 2010, 24(4):429443.
［23］	Narang H,Wu F,Cabral M.Numerical solutions of heat and mass transfer in capillary porous media using programmable graphics hardware［M］∥Recent Advances in Computer Science and Information Engineering, Berlin:Springer Berlin Heidelberg, 2012:127134.
［24］	Liu Xin. Parallel computing and largescale parallel computing platform of chemical nonequilibrium floworiented CFD［D］. Zhenzhou:PLA Information Engineering University, 2006.(in Chinese)
［25］	Ding Yubo, Chen Xinhua, Wang Hai. Multidomain parallel computing for static strength analysis of whole aircraft model［J］. Aeronautical Computing Technique, 2010, 40(5):6769.(in Chinese)
［26］	Ruan Honghe, Yuan Yong, Liu Xian. Progress of parallel finite element method on distributedmemory parallel computer systems［J］. Journal of Tongji University(Natural Science), 2005, 33(1):2127.(in Chinese)
［27］	Zheng G, Wilmarth T, Lawlor O, et al. Performance modeling and programming environments for petaflops computers and the blue gene machine［C］∥Proc of the 18th International Parallel and Distributed Processing Symposium, 2004:197.
［28］	Ghosh D, Avery P, Farhat C. A FETIpreconditioned conjugate gradient method for largescale stochastic finite element problems［J］. International Journal for Numerical Methods in Engineering, 2009, 80(67):914931.
［29］	Xia Yidong, Wu Yizhao, Lv Hongqiang. et al. Parallel computation of a highorder discontinuous Galerkin method on unstructured grids［J］. Acta Aerodynamica Sinica, 2011, 29(5):537541.(in Chinese)
［30］	Liu Y, Jiao S, Wu W, et al. GPU accelerated fast FEM deformation simulation［C］∥IEEE Asia Pacific Conference on Circuits and Systems,2008:606609.
［31］	Singh I, Jain P. Parallel EFG algorithm for heat transfer problems［J］. Advances in Engineering Software, 2005, 36(8):554560.
［32］	Ltaief H, Gabriel E, Garbey M. Fault tolerant algorithms for heat transfer problems［J］. Journal of Parallel and Distributed Computing, 2008, 68(5):663677.
［33］	Mohammadzadeh A, Roohi E, Niazmand H, et al. Thermal and secondlaw analysis of a microor nanocavity using directsimulation Monte Carlo［J］. Physical Review E, 2012, 85(5):056310.
［34］	Obrecht C, Kuznik F, Tourancheau B, et al. MultiGPU implementation of a hybrid thermal lattice Boltzmann solver using the TheLMA framework［J］. Computers & Fluids, 2013,80(10):269275.
［35］	Obrecht C, Kuznik F, Tourancheau B, et al. The TheLMA project:A thermal lattice Boltzmann solver for the GPU［J］. Computers & Fluids. 2012, 54(0):118126.
［36］	Fu You, Hua Rong, Kang Jichang. Migration dependency analysis of DSMC parallel simulation［J］. Microelectronics & Computer, 2007, 24(5):175178.(in Chinese)
［37］	Wang Xuede, Wu Yizhao, Xia Jian. A parallel algorithm of 2D unstructured DSMC method with dynamic load balance［J］. ACTA Aerodynamica SINICA, 2007, 25(3):339344.(in Chinese)
［38］	Ivanov M, Markelov G, Taylor S, et al. Parallel DSMC strategies for 3D computations［C］∥Proc of Parallel CFD, 1996:485492.
［39］	Scanlon T, Roohi E, White C, et al. An open source, parallel DSMC code for rarefied gas flows in arbitrary geometries［J］. Computers & Fluids, 2010, 39(10):20782089.
［40］	Xiao Honglin, Wang Liansheng. Investigation on parallel computing algorithm of DNS using pseudospectral method based on MPI［J］. Computer Engineering and Applications, 2012, 48(4):5457.(in Chinese)
［41］	Oak Ridge National Laboratory(ORNL). Oak ridge leadership computing facility［EB/OL］.［20130301］.http://www.olcf.ornl.gov/leadershipscience/engineering/.
［42］	Chen Y, Cui X, Mei H. PARRAY:A unifying array representation for heterogeneous parallelism［C］∥Proc of the 17th ACM SIGPLAN symposium on Principles and Practice of Parallel Programming, 2012:171180.
［43］	Gong Chunye,Liu Jie,Huang Haowei,et al.Particle transport with unstructured grid on GPU［J］. Computer Physics Communications,2012,183(3):588593.
［44］	Gong Chunye, Liu Jie, Chi Lihua, et al. GPU accelerated simulations of 3D deterministic particle transport using discrete ordinates method［J］. Journal of Computational Physics, 2011, 230(15):6010  6022.
［45］	Mo Zeyao, Zhang Aiqing, Cao Xiaolin, et al. JASMIN:A parallel software infrastructure for scientific computing［J］. Frontiers of Computer Science in China, 2010,4(4):480488.
	附中文参考文献：
［3］	张林波, 迟学斌, 莫则尧, 李若. 并行计算导论［M］. 北京:清华大学出版社, 2006.
［6］	刘晨. 复杂燃烧流场数值模拟方法研究［D］. 南京:南京航空航天大学, 2009.
［7］	中国中央政府. 中华人民共和国国民经济和社会发展第十二个五年规划纲要［EB/OL］.［20130301］.http://www.gov.cn/.
［8］	中国国家自然科学基金委员会. 国家自然科学基金“十二五” 发展规则［EB/OL］.［20130301］.http://www.nsfc.gov.cn/.
［9］	中国国家科技部. 关于征集十二五863计划高效能计算机及应用服务环境重大项目课题建议的通知［EB/OL］.［20130301］.http://www.most.gov.cn/.
［11］	都志辉. 高性能计算之并行编程技术——MPI 并行程序设计［M］.北京:清华大学出版社, 2001.
［13］	张涵信, 沈孟育. 计算流体力学:差分方法的原理和应用［M］. 北京:国防工业出版社,2003.
［21］	潘沙. 高超声速气动热数值模拟方法及大规模并行计算研究［D］. 长沙:国防科学技术大学, 2010.
［24］	刘鑫. 面向化学非平衡流的CFD 并行计算技术和大规模并行计算平台研究［D］. 郑州:解放军信息工程大学, 2006.
［25］	丁玉波, 陈秀华, 汪海. 全机模型静强度分析的多区域并行计算［J］. 航空计算技术,2010, 40(5):6769.
［26］	阮红河, 袁勇, 柳献. 分布存储环境并行有限元研究进展［J］. 同济大学学报:自然科学版,2005, 33(1):2127.
［29］	夏轶栋, 伍贻兆, 吕宏强, 等. 高阶间断有限元法的并行计算研究［J］. 空气动力学学报,2011, 29(5):537541.
［36］	傅游, 花嵘, 康继昌. DSMC并行仿真中的迁移相关分析方法［J］. 微电子学与计算机,2007, 24(5):175178.
［37］	王学德, 伍贻兆, 夏健. 动态负载平衡的二维非结构网格DSMC 并行算法研究［J］. 空气动力学学报,2007, 25(3):339344.
［40］	肖红林, 王连生. 基于MPI 的伪谱法DNS 并行计算方法研究［J］. 计算机工程与应用,2012, 48(4):5457.

[1]	刘屹成, 刘晓燕, 严馨. 并行平衡级联支持向量机[J]. 计算机工程与科学, 2023, 45(07): 1170-1177.
[2]	吴铁彬, 过锋, 王谛. 面向E级计算的高性能处理器核心运算架构研究进展[J]. 计算机工程与科学, 2023, 45(05): 761-771.
[3]	臧照虎, 李晨, 王耀华, 陈小文, 郭阳. 面向众核系统的层次化栅栏同步机制[J]. 计算机工程与科学, 2022, 44(11): 1901-1908.
[4]	张勇, 张曦, 万云博, 何先耀, 赵钟, 卢宇彤. 非结构有限体积CFD计算的网格重排序优化[J]. 计算机工程与科学, 2022, 44(10): 1721-1729.
[5]	陈奉贤. 基于NR-Transformer的集群作业运行时间预测[J]. 计算机工程与科学, 2022, 44(07): 1181-1190.
[6]	曹继军. 面向HPC和DC的可重构光互连网络体系结构综述[J]. 计算机工程与科学, 2022, 44(06): 951-963.
[7]	袁国兴, 张云泉, 袁良. 2021年中国高性能计算机发展现状分析[J]. 计算机工程与科学, 2021, 43(12): 2091-2097.
[8]	范培勤, 过武宏, 韩梅, 唐帅, 张驰, . 水声环境特征参数并行预报方法研究[J]. 计算机工程与科学, 2021, 43(11): 1920-1925.
[9]	袁远, 李世杰, 邢建英, 蒋句平. E级高性能计算机系统中监控分系统的挑战与设计[J]. 计算机工程与科学, 2021, 43(08): 1366-1375.
[10]	龚昊, 刘莹, 冯建周, 赵仁良, 冷佳旭, . 基于GPU加速的脉冲多普勒雷达信号处理[J]. 计算机工程与科学, 2021, 43(07): 1141-1149.
[11]	焦育威, 王鹏, 辛罡, . 基于采样尺度自适应的多尺度量子谐振子优化算法并行化[J]. 计算机工程与科学, 2021, 43(07): 1200-1209.
[12]	俞茂学, 贾东宁, 魏志强, 许佳立, 马广浩. 一种基于国产异构众核处理器的C++智能源码转换框架[J]. 计算机工程与科学, 2021, 43(06): 997-1005.
[13]	陈云, 王梦园, 柴晓楠, 商建东, . 面向FT-M7002的高斯滤波算法优化实现[J]. 计算机工程与科学, 2021, 43(05): 799-806.
[14]	丁哲昭, 储根深, 胡长军, 李扬. 基于申威众核处理器的圣维南求解程序的并行与优化[J]. 计算机工程与科学, 2021, 43(05): 820-829.
[15]	赵永浩, 贾海鹏, 张云泉, 张思佳. 基于SIMD的Square Root函数高性能实现与优化[J]. 计算机工程与科学, 2021, 43(04): 662-669.