29-Billion Atoms Molecular Dynamics Simulation With <i>Ab Initio</i> Accuracy on 35 Million Cores of New Sunway Supercomputer

doi:10.1109/TC.2025.3540646

	*29-Billion Atoms Molecular Dynamics Simulation With Ab Initio* Accuracy on 35 Million Cores of New Sunway Supercomputer**
	Wang, Xun 1; Meng, Xiangyu 1; Guo, Zhuoqiang 2,3; Li, Mingzhen 2,3; Liu, Lijun 4; Li, Mingfan 5; Xiao, Qian 5; Zhao, Tong 2,3; Sun, Ninghui 2,3; Tan, Guangming 2,3; Jia, Weile 2,3
	2025-05-01
发表期刊	IEEE TRANSACTIONS ON COMPUTERS
ISSN	0018-9340
卷号	74 期号:5 页码:1634-1648
摘要	Physical phenomena such as bond breaking and phase transitions require molecular dynamics (MD) with ab initio accuracy, involving up to billions of atoms and over nanosecond timescales. Previous state-of-the-art work has demonstrated that neural network molecular dynamics (NNMD) like deep potential molecular dynamics (DeePMD), can successfully extend the temporal and spatial scales of MD with ab initio accuracy on both ARM and GPU platforms. However, the DeePMD-kit package is currently unable to fully exploit the computational potential of the new Sunway supercomputer due to its unique many-core architecture, memory hierarchy, and low precision capability. In this paper, we re-design the DeePMD-kit to harness the massive computing power of the new Sunway, enabling the MD with over ten billion atoms. We first design a large-scale parallelization scheme to exploit the massive parallelism of the new Sunway. Then we devise specialized optimizations for the time-consuming operators. Finally, we design a novel mixed precision method for DeePMD-kit customized operators to leverage the low precision computing power of the new Sunway. The optimized DeePMDkit achieves 67.6 / 56.5 x speedup for water / copper systems on the new Sunway. Meanwhile, it can perform 29 billion atoms simulation for the water system on 35 million cores (i.e., 90,000 computing nodes, around 84% of the whole supercomputer) with a peak performance of 57.1 PFLOPs, which is 7.9x bigger and 1.2x faster than state-of-the-art results. This paves the way for investigating more realistic scenarios, such as studying the mechanical properties of metals, semiconductor devices, batteries, and other materials and physical systems.
关键词	Atoms Accuracy Supercomputers Optimization Artificial neural networks Force Training Fitting Predictive models Nuclear power generation High Performance Computing Molecular Dynamics DeePMD Parallel Optimization New Sunway Supercomputer
DOI	10.1109/TC.2025.3540646
收录类别	SCI
语种	英语
资助项目	National Science Foundation of China[62372435] ; National Science Foundation of China[T2125013] ; National Science Foundation of China[62032023] ; National Science Foundation of China[61972377] ; National Science Foundation of China[92270206] ; National Science Foundation of China[61972380] ; National Science Foundation of China[61972416] ; CAS Project for Young Scientists in Basic Research[YSBR-005] ; China National Postdoctoral Program for Innovative Talents[BX20240383] ; Natural Science Foundation of Shandong Province[ZR2022LZH009] ; GHfund C[202407035455] ; National Key R&D Program of China[2021YFA1000103-3]
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Hardware & Architecture ; Engineering, Electrical & Electronic
WOS记录号	WOS:001463492900015
出版者	IEEE COMPUTER SOC
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/40587
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Wang, Xun; Li, Mingzhen; Jia, Weile
作者单位	1.China Univ Petr East China, Dept Comp Sci & Technol, Shandong Key Lab Intelligent Oil & Gas Ind Softwar, Qingdao 266580, Shandong, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, State Key Lab Proc, Beijing 100190, Peoples R China 3.Univ Chinese Acad Sci, Beijing 100190, Peoples R China 4.Osaka Univ, Dept Mech Engn, Suita, Osaka 5650871, Japan 5.Univ Sci & Technol China, Sch Comp Sci & Technol, Hefei 230027, Peoples R China
推荐引用方式 GB/T 7714	Wang, Xun,Meng, Xiangyu,Guo, Zhuoqiang,et al. 29-Billion Atoms Molecular Dynamics Simulation With Ab Initio Accuracy on 35 Million Cores of New Sunway Supercomputer[J]. IEEE TRANSACTIONS ON COMPUTERS,2025,74(5):1634-1648.
APA	Wang, Xun.,Meng, Xiangyu.,Guo, Zhuoqiang.,Li, Mingzhen.,Liu, Lijun.,...&Jia, Weile.(2025).29-Billion Atoms Molecular Dynamics Simulation With Ab Initio Accuracy on 35 Million Cores of New Sunway Supercomputer.IEEE TRANSACTIONS ON COMPUTERS,74(5),1634-1648.
MLA	Wang, Xun,et al."29-Billion Atoms Molecular Dynamics Simulation With Ab Initio Accuracy on 35 Million Cores of New Sunway Supercomputer".IEEE TRANSACTIONS ON COMPUTERS 74.5(2025):1634-1648.