We use cookies to improve your experience with our site.
Yu-Jin Yan, Hai-Bo Li, Tong Zhao, Lin-Wang Wang, Lin Shi, Tao Liu, Guang-Ming Tan, Wei-Le Jia, Ning-Hui Sun. 10-million atoms simulation of first-principle package LS3DF on Sugon supercomputer[J]. Journal of Computer Science and Technology. DOI: 10.1007/s11390-023-3011-6
Citation: Yu-Jin Yan, Hai-Bo Li, Tong Zhao, Lin-Wang Wang, Lin Shi, Tao Liu, Guang-Ming Tan, Wei-Le Jia, Ning-Hui Sun. 10-million atoms simulation of first-principle package LS3DF on Sugon supercomputer[J]. Journal of Computer Science and Technology. DOI: 10.1007/s11390-023-3011-6

10-million atoms simulation of first-principle package LS3DF on Sugon supercomputer

  • The growing demand for semiconductor devices simulation poses a big challenge for large-scale electronic structure calculations. Among various methods, the linear scaling three dimensional fragment method (LS3DF) 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 the Sugon supercomputer, a domestic supercomputer equipped with Deep Computing Units (DCU). 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 10-million silicon 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.
  • loading

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return