PROJECT 1Development of GENESIS for high performance MD

Many biologically important processes occur on the time scales of microseconds, milliseconds or even larger while the used time step in MD is one or two femtoseconds. Therefore, in MD, there are inherent limitations in simulation time and target system size. We have developed GENESIS with efficient parallelization and have succeeded MD simulations of systems consisting of 100 million ~ 1 billion atom systems. Based on the development of GENESIS with efficient parallelization, we further increase the performance from optimizations for next-generation supercomputers and new hardware architectures including GPUs and coprocessors as well as CPUs. To understand biological phenomena at very large scale, we develop high performance schemes for various coarse-grained levels. We also develop efficient integration schemes enabling large time step without any loss of accuracy. We expect that these developments make a new breakthrough that overcome the current limitations in MD.

References

  1. “New parallel computing algorithm of molecular dynamics for extremely huge scale biological systems.”, Jaewoon Jung, Chigusa Kobayashi, Kento Kasahara, Cheng Tan, Akiyoshi Kuroda, Kazuo Minami, Shigeru Ishiduki, Tatsuo Nishiki, Hikaru Inoue, Yutaka Ishikawa, Michael Feig, and Yuji Sugita, J. Comput. Chem, 42, 231-241 (2021).
  2. “Group-based evaluation of temperature and pressure for molecular dynamics simulation with a large time steps.”, Jaewoon Jung and Yuji Sugita, J. Chem. Phys, 153, 234115 (2020).
  3. “Scaling molecular dynamics beyond 100,000 processor cores for large-scale biophysical simulations”, Jaewoon Jung, Wataru Nishima, Marcus Daniels, Gavin Bascom, Chigusa Kobayashi, Adetokunbo Adedoyin, Michael Wall, Anna Lappala, Dominic Phillips, William Fischer, Chang-Shung Tung, Tamar Schlick, Yuji Sugita, and Karissa Y. Sanbonmatsu., J. Comput. Chem, 40, 1919-1930 (2019).
  4. “Optimal temperature evaluation in molecular dynamics simulations with a large time step” J. Jung, C. Kobayashi, and Y. Sugita, J. Chem. Theory Comput., 15, 84-94 (2019).
  5. “GENESIS 1.1: A hybrid-parallel molecular dynamics simulator with enhanced sampling algorithms on multiple computational platforms”, Chigusa Kobayashi, Jaewoon Jung, Yasuhiro Matsunaga, Takaharu Mori, Tadashi Ando, Koichi Tamura, Motoshi Kamiya, and Yuji Sugita., J. Comput. Chem., 38, 2193-2206 (2017).
  6. “Parallel implementation of 3D FFT with volumetric decomposition schemes for efficient molecular dynamics simulations.” Jaewoon Jung, Chigusa Kobayashi, Toshiyuki Imamura, and Yuji Sugita, Comput. Phys. Comm., 200, 57-65 (2016).
  7. “Graphics processing unit acceleration and parallelization of GENESIS for large-scale molecular dynamics simulations”, Jaewoon Jung, Akira Naruse, Chigusa Kobayashi, and Yuji Sugita., J. Chem. Theory Comput., 12, 4947–4958 (2016).
  8. “GENESIS: a hybrid-parallel and multi-scale molecular dynamics simulator with enhanced sampling algorithms for biomolecular and cellular simulations”, Jaewoon Jung, Takaharu Mori, Chigusa Kobayashi, Yasuhiro Matsunaga, Takao Yoda, Michael Feig, and Yuji Sugita, WIREs Comput. Mol. Sci., 5, 310–323 (2015).