基于神威太湖之光的宇宙学多体模拟

doi:10.19678/j.issn.1000-3428.0056967

计算机工程 ›› 2020, Vol. 46 ›› Issue (9): 35-43. doi: 10.19678/j.issn.1000-3428.0056967

基于神威太湖之光的宇宙学多体模拟

刘旭^1,2, 张曦煌¹, 刘钊^2,3, 吕小敬², 朱光辉⁴

1. 江南大学物联网工程学院, 江苏无锡 214122;
2. 国家超级计算无锡中心, 江苏无锡 214072;
3. 清华大学, 北京 100084;
4. 无锡航天江南数据系统科技有限公司, 江苏无锡 214000

收稿日期:2019-12-19 修回日期:2020-02-28 发布日期:2020-03-06
作者简介:刘旭(1995-),男,硕士研究生,主研方向为高性能计算、并行编译及优化;张曦煌,教授、博士;刘钊,工程师、博士研究生;吕小敬、朱光辉,工程师、硕士。
基金资助:
国家自然科学基金（51877115）。

Cosmological Multi-Body Simulation Based on Sunway TaihuLight

LIU Xu^1,2, ZHANG Xihuang¹, LIU Zhao^2,3, LÜ Xiaojing², ZHU Guanghui⁴

1. School of Internet of Things Engineering, Jiangnan University, Wuxi, Jiangsu 214122, China;
2. National Supercomputing Center in Wuxi, Wuxi, Jiangsu 214072, China;
3. Tsinghua University, Beijing 100084, China;
4. Wuxi Aerospace Jiangnan Data System Technology Co., Ltd., Wuxi, Jiangsu 214000, China

Received:2019-12-19 Revised:2020-02-28 Published:2020-03-06

摘要/Abstract

摘要： 宇宙学模拟对于科学家研究非线性结构的形成以及暗物质、暗能量等假想形式具有重要作用，而高精度宇宙学模拟包含数千亿甚至数万亿个粒子，因此超级计算机强大的计算能力使其成为解决宇宙学模拟问题的理想平台。为在国产神威太湖之光超级计算机上实现宇宙学N体模拟，分析PHoToNs软件中使用的粒子网格算法和快速多极子方法，并结合众核处理器架构提出多层次分解和负载均衡方案、执行树遍历和引力计算的流水线策略以及向量化引力计算算法等多种性能优化技术，从而实现能充分发挥神威太湖之光架构优势的N体模拟软件SwPHoToNs。实验结果表明，在神威太湖之光超级计算系统的5 200 000个计算核心上进行包含6 400亿个粒子的宇宙学模拟，SwPHoToNs获得了29.44 PFLOPS的持续计算速度，且并行和计算效率分别为84.6%和48.3%。

关键词: 神威太湖之光, 宇宙学, 多体模拟, 并行优化, 可扩展性

Abstract: Cosmological simulations are essential for scientists to study the formation of non-linear structures and hypotheses of dark matter,dark energy,etc.High-precision cosmological simulations include hundreds of billions or even trillions of particles,thus demanding massive computational power.So supercomputers can provide an ideal platform for cosmological simulation.To implement cosmological N-body simulation on Sunway TaihuLight,a supercomputer developed in China,this paper analyzes the Particle Mesh(PM) and Fast Multipole Method(FMM) in PHoToNs.The analysis results are combined with the multi-core processor structure,and on this basis this paper proposes multiple performance optimization techniques,including a multi-level decomposition and load balancing scheme,a pipeline strategy using execution tree traversal and gravity calculation,and a vectorized gravity calculation algorithm.By using the above techniques,a N-body simulation software,SwPHoToNs,is implemented,which can give full play to the structural advantages of Sunway TaihuLight.Experimental results show that when conducting cosmological simulations which contain up to 640 billion particles on 5 200 000 cores of Sunway TaihuLight,SwPHoToNs obtains a sustained calculation speed of 29.44 PFLOPS with a parallel efficiency of 84.6% and computational efficiency of 48.3%.

Key words: Sunway TaihuLight, cosmology, multi-body simulation, parallel optimization, scalability

中图分类号:

TP399

刘旭, 张曦煌, 刘钊, 吕小敬, 朱光辉. 基于神威太湖之光的宇宙学多体模拟[J]. 计算机工程, 2020, 46(9): 35-43.

LIU Xu, ZHANG Xihuang, LIU Zhao, LÜ Xiaojing, ZHU Guanghui. Cosmological Multi-Body Simulation Based on Sunway TaihuLight[J]. Computer Engineering, 2020, 46(9): 35-43.

https://www.ecice06.com/CN/Y2020/V46/I9/35

图/表 11

20200915134309

20200915134332

20200915134335

20200915134341

20200915134348

20200915134351

20200915134418

20200915134422

20200915134426

20200915134429

20200915134432

参考文献

[1] WANG Qiao,CAO Zongyan,GAO Liang,et al.PHoToNs-a parallel heterogeneous and threads oriented code for cosmological N-body simulation[J].Research in Astronomy and Astrophysics,2018,18(6):23-62.
[2] BARNES J,HUT P.A hierarchical O(NlogN) force-calculation algorithm[J].Nature,1986,324(6096):446-449.
[3] HOCKNEY R,EASTWOOD J.Computer simulation using particles[M].[S.l.]:Taylor & Francis,Inc.,1988.
[4] GREENGARD L,ROKHLIN V.A fast algorithm for particle simulations[J].Journal of Computational Physics,1987,73(2):325-348.
[5] WARREN M S,SALMON J K.Astrophysical N-body simulations using hierarchical tree data structures[C]//Proceedings of 1992 ACM/IEEE Conference on Super-computing.Washington D.C.,USA:IEEE Press,1992:570-576.
[6] WARREN M S.2HOT:an improved parallel hashed OCT-tree N-body algorithm for cosmological simulation[J].Scientific Programming,2014,22(2):109-124.
[7] ISHIYAMA T,NITADORI K,MAKINO J.4.45 Pflops astrophysical N-body simulation on K computer-the gravitational trillion-body problem[C]//Proceedings of International Conference on High Performance Computing,Networking,Storage and Analysis.Washington D.C.,USA:IEEE Press,2012:1-10.
[8] MAKINO J,TAIJI M.Astrophysical N-body simulations on GRAPE-4 special-purpose computer[C]//Proceedings of 1995 ACM/IEEE Conference on Supercomputing.Washington D.C.,USA:IEEE Press,1995:63-75.
[9] KAWAI A,FUKUSHIGE T,MAKINO J.$7.0/Mflops astrophysical N-body simulation with treecode on GRAPE-5[C]//Proceedings of 1999 ACM/IEEE Conference on Supercomputing.Washington D.C.,USA:IEEE Press,1999:67-85.
[10] MAKINO J,FUKUSHIGE T,KOGA M.A 1.349 Tflops simulation of black holes in a galactic center on GRAPE-6[C]//Proceedings of 2000 ACM/IEEE Conference on Supercomputing.Washington D.C.,USA:IEEE Press,2000:43-53.
[11] MAKINO J,KOKUBO E,FUKUSHIGE T.Performance evaluation and tuning of GRAPE-6-towards 40"real" TFlops[C]//Proceedings of 2003 ACM/IEEE Conference on Supercomputing.Washington D.C.,USA:IEEE Press,2003:22-116.
[12] HAMADA T,NARUMI T,YOKOTA R,et al.42 TFlops hierarchical N-body simulations on GPUs with applications in both astrophysics and turbulence[C]//Proceedings of Conference on High Performance Computing Networking,Storage and Analysis.Washington D.C.,USA:IEEE Press,2009:1-12.
[13] HAMADA T,NITADORI K.190 TFlops astrophysical N-body simulation on a cluster of GPUs[C]//Proceedings of 2010 ACM/IEEE International Conference for High Performance Computing,Networking,Storage and Analysis.Washington D.C.,USA:IEEE Press,2010:1-9.
[14] BÉDORF J,GABUROV E,FUJII M S,et al.24.77 Pflops on a gravitational tree-code to simulate the Milky way galaxy with 18600 GPUs[C]//Proceedings of International Conference for High Performance Computing,Networking,Storage and Analysis.Washington D.C.,USA:IEEE Press,2014:54-65.
[15] IWASAWA M,WANG L,NITADORI K,et al.Global simulation of planetary rings on Sunway TaihuLight[C]//Proceedings of International Conference on Computational Science.Berlin,Germany:Springer,2018:483-495.
[16] FANG Jiarui,FU Haohuan,ZHAO Wenlai,et al.swDNN:a library for accelerating deep learning applications on Sunway TaihuLight[C]//Proceedings of 2017 IEEE International Parallel and Distributed Processing Symposium.Washington D.C.,USA:IEEE Press,2017:615-624.
[17] XU LEI,XU YING.Investigation of the implementation of N-body problem on GPUs[J].Computer Applications and Software,2012,29(1):92-95.(in Chinese)徐磊,徐莹.多体问题在GPU上实现的讨论[J].计算机应用与软件,2012,29(1):92-95.
[18] LI Ning.2DECOMP&FFT-a highly scalable 2D decomposi-tion library and FFT interface[C]//Proceedings of CUG'10.Washington D.C.,USA:IEEE Press,2010:24-27.
[19] DEHNEN W.A hierarchical O(N) force calculation algorithm[J].Journal of Computational Physics,2002,179(1):27-42.
[20] PRUNET S,PICHON C,AUBERT D,et al.Initial conditions for large cosmological simulations[J].The Astrophysical Journal Letters Supplement Series,2008,178(2):179-188.

选择文件类型/文献管理软件名称

选择包含的内容

基于神威太湖之光的宇宙学多体模拟

Cosmological Multi-Body Simulation Based on Sunway TaihuLight

RichHTML

PDF

可视化

被引次数

摘要/Abstract

引用本文

使用本文

图/表 11

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	王华维, 刘若妍, 艾志玮, 曹轶. 基于多绘制管线的大规模并行体绘制性能优化技术[J]. 计算机工程, 2024, 50(8): 207-215.
[2]	黄斌, 柳安军, 潘景山, 田敏, 张煜, 朱光慧. 基于GPU的LBM迁移模块算法优化[J]. 计算机工程, 2024, 50(2): 232-238.
[3]	苏瑞国, 阳建, 秦继伟, 武晓雄, 贾振红. 基于物联网区块链的轻量级共识算法研究[J]. 计算机工程, 2023, 49(2): 175-180.
[4]	黄华威, 孔伟, 彭肖文, 郑子彬. 区块链分片技术综述[J]. 计算机工程, 2022, 48(6): 1-10.
[5]	彭龙, 陈俊仕, 安虹. 基于神威太湖之光的AMBER软件移植与优化[J]. 计算机工程, 2020, 46(12): 12-20.
[6]	彭振,吴百锋. 基于数据并行的碰撞检测[J]. 计算机工程, 2017, 43(9): 1-6.
[7]	陈曦,朱建涛,何晓斌. 一种面向高性能计算的分布式对象存储系统[J]. 计算机工程, 2017, 43(8): 69-73.
[8]	张剑,刘梦赤. 面向信息网模型的高可扩展性分布式事务处理机制[J]. 计算机工程, 2017, 43(11): 16-21,26.
[9]	马冬冬,衷璐洁,朱敬茹. 基于GPU的LLVM程序分析信息并行提取[J]. 计算机工程, 2017, 43(10): 23-30.
[10]	蒋成. 基于弧形不相交生成树的多链路故障恢复方法[J]. 计算机工程, 2016, 42(7): 65-71.
[11]	周文剑,靳婷. Blogel-SPINAL:分布式PPI网络比对算法[J]. 计算机工程, 2016, 42(4): 209-214,220.
[12]	陈振武,郑汉垣,兰添才,曾志宏. 求解大规模三对角线性方程组的GaBP并行算法[J]. 计算机工程, 2016, 42(10): 96-100.
[13]	胡新明，盛冲冲，李佳佳，吴百锋. 面向通用计算GPU集群的任务自动分配系统[J]. 计算机工程, 2014, 40(3): 103-107,119.
[14]	刘石柱, 尹首一, 殷崇勇, 刘雷波, 魏少军. 基于可重构处理器的并行优化算法[J]. 计算机工程, 2012, 38(21): 286-289.
[15]	陈庆奎, 曹欢欢, 庄松林. 一种GPU集群的动态任务映射策略[J]. 计算机工程, 2012, 38(17): 268-271.

模态框（Modal）标题

选择文件类型/文献管理软件名称

选择包含的内容

基于神威太湖之光的宇宙学多体模拟

Cosmological Multi-Body Simulation Based on Sunway TaihuLight

RichHTML

PDF

可视化

被引次数

摘要/Abstract

引用本文

使用本文

图/表 11

参考文献

相关文章 15

编辑推荐

Metrics

本文评价