Institute of Computing Technology, Chinese Academy IR
10-Million Atoms Simulation of First-Principle Package LS3DF | |
Yan, Yu-Jin1,2; Li, Hai-Bo1,3; Zhao, Tong2; Wang, Lin-Wang4; Shi, Lin5; Liu, Tao1; Tan, Guang-Ming1; Jia, Wei-Le1; Sun, Ning-Hui1,2 | |
2024-02-01 | |
发表期刊 | JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY |
ISSN | 1000-9000 |
卷号 | 39期号:1页码:45-62 |
摘要 | The growing demand for semiconductor devices simulation poses a big challenge for large-scale electronic structure calculations. Among various methods, the linearly scaling three-dimensional fragment (LS3DF) method exhibits excellent scalability in large-scale simulations. Based on algorithmic and system-level optimizations, we propose a highly scalable and highly efficient implementation of LS3DF on a domestic heterogeneous supercomputer equipped with accelerators. In terms of algorithmic optimizations, the original all-band conjugate gradient algorithm is refined to achieve faster convergence, and mixed precision computing is adopted to increase overall efficiency. In terms of system-level optimizations, the original two-layer parallel structure is replaced by a coarse-grained parallel method. Optimization strategies such as multi-stream, kernel fusion, and redundant computation removal are proposed to increase further utilization of the computational power provided by the heterogeneous machines. As a result, our optimized LS3DF can scale to a 10-million silicon atoms system, attaining a peak performance of 34.8 PFLOPS (21.2% of the peak). All the improvements can be adapted to the next-generation supercomputers for larger simulations. |
关键词 | single instruction multiple thread accelerator electronic structure high-performance computing linearly scaling three-dimensional fragment (LS3DF) |
DOI | 10.1007/s11390-023-3011-6 |
收录类别 | SCI |
语种 | 英语 |
资助项目 | National Key Research and Development Program of China[2021YFB0300600] ; National Natural Science Foundation of China[92270206] ; National Natural Science Foundation of China[T2125013] ; National Natural Science Foundation of China[62032023] ; National Natural Science Foundation of China[61972377] ; National Natural Science Foundation of China[T2293702] ; National Natural Science Foundation of China[12274360] ; Chinese Academy of Sciences Project for Young Scientists in Basic Research[YSBR-005] ; Network Information Project of Chinese Academy of Sciences[CASWX2021SF-0103] ; Key Research Program of Chinese Academy of Sciences[ZDBSSSW-WHC002] |
WOS研究方向 | Computer Science |
WOS类目 | Computer Science, Hardware & Architecture ; Computer Science, Software Engineering |
WOS记录号 | WOS:001200760400014 |
出版者 | SPRINGER SINGAPORE PTE LTD |
引用统计 | |
文献类型 | 期刊论文 |
条目标识符 | http://119.78.100.204/handle/2XEOYT63/38983 |
专题 | 中国科学院计算技术研究所期刊论文_英文 |
通讯作者 | Li, Hai-Bo; Jia, Wei-Le; Sun, Ning-Hui |
作者单位 | 1.Chinese Acad Sci, Inst Comp Technol, State Key Lab Proc, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Beijing 101408, Peoples R China 3.Huawei Technol, Comp Syst Optimizat Lab, Beijing 100094, Peoples R China 4.Chinese Acad Sci, Inst Semicond, Beijing 100083, Peoples R China 5.Yancheng Inst Technol, Sch Mat Sci & Engn, Yancheng 224051, Peoples R China |
推荐引用方式 GB/T 7714 | Yan, Yu-Jin,Li, Hai-Bo,Zhao, Tong,et al. 10-Million Atoms Simulation of First-Principle Package LS3DF[J]. JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY,2024,39(1):45-62. |
APA | Yan, Yu-Jin.,Li, Hai-Bo.,Zhao, Tong.,Wang, Lin-Wang.,Shi, Lin.,...&Sun, Ning-Hui.(2024).10-Million Atoms Simulation of First-Principle Package LS3DF.JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY,39(1),45-62. |
MLA | Yan, Yu-Jin,et al."10-Million Atoms Simulation of First-Principle Package LS3DF".JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY 39.1(2024):45-62. |
条目包含的文件 | 条目无相关文件。 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论