Tree cutting approach for domain partitioning on forest-of-octrees-based block-structured static adaptive mesh refinement with lattice Boltzmann method

Hasegawa, Yuta; Aoki, Takayuki*; Kobayashi, Hiromichi*; Idomura, Yasuhiro; Onodera, Naoyuki

Parallel Computing, 108, p.102851_1 - 102851_12, 2021/12

https://doi.org/10.1016/j.parco.2021.102851

The aerodynamics simulation code based on the lattice Boltzmann method (LBM) using forest-of-octrees-based block-structured local mesh refinement (LMR) was implemented, and its performance was evaluated on GPU-based supercomputers. We found that the conventional Space-Filling-Curve-based (SFC) domain partitioning algorithm results in costly halo communication in our aerodynamics simulations. Our new tree cutting approach improved the locality and the topology of the partitioned sub-domains and reduced the communication cost to one-third or one-fourth of the original SFC approach. In the strong scaling test, the code achieved maximum speedup at the performance of 2207 MLUPS (mega- lattice update per second) on 128 GPUs. In the weak scaling test, the code achieved 9620 MLUPS at 128 GPUs with 4.473 billion grid points, while the parallel efficiency was 93.4% from 8 to 128 GPUs.

Development of production methods of the Sr-90/Y-90 source for hydrogen production experiments

Motoki, Riyozo; Aoki, Hiromichi; Uchida, Shoji; Nagaishi, Ryuji; Yamada, Reiji

JAEA-Technology 2008-014, 23 Pages, 2008/03

JAEA-Technology-2008-014.pdf:9.05MB

The study of producing hydrogen with a Sr-90/Y-90 source is planned to utilze the radioactive waste effectively. Therefore we developed two methods of caking Sr-90 and a catalyst for the production of hydrogen effectively. One is a method of caking SrTiO and TiO in a silica gel. And another is a method of caking SrSO and TiO in a silica gel. These solid matters are porous materials, which has a radiation resistant and chemical resistant. In addition, Y-90 which is a daughter nuclide of Sr-90 can be also used for hydrogen production.

Analytical work at NUCEF in FY 2005

Fukaya, Hiroyuki; Aoki, Hiromichi; Haga, Takahisa; Nishizawa, Hidetoshi; Sonoda, Takashi; Sakazume, Yoshinori; Shimizu, Kaori; Niitsuma, Yasushi*; Shirahashi, Koichi; Inoue, Takeshi

JAEA-Technology 2007-005, 27 Pages, 2007/03

JAEA-Technology-2007-005.pdf:1.97MB

Analysis of the uranyl nitrate solution fuel is carried out at the analytical laboratory of NUCEF (Nuclear Fuel Cycle Engineering Research Facility), which provides essential data for operation of STACY (Static Experiment Critical Facility), TRACY (Transient Experiment Critical Facility) and the fuel treatment system. Analyzed in FY 2005 were uranyl nitrate solution fuel samples taken before and after experiments of STACY and TRACY, samples for the preparation of uranyl nitrate solution fuel, and samples for nuclear material accountancy purpose. Also analyzed were the samples from raffinate treatment and its preliminary tests. The raffinate was generated, since FY 2000, during preliminary experiments on U/Pu extraction-pulification to fix the operation condition to prepare plutonium solution fuel to be used for criticality experiments at STACY. This report summarizes work related to the analysis and management of the analytical laboratory in the FY 2005.

Tree cutting approach for reducing communication in domain partitioning of tree-based block-structured adaptive mesh refinement

Hasegawa, Yuta; Aoki, Takayuki*; Kobayashi, Hiromichi*; Idomura, Yasuhiro; Onodera, Naoyuki

no journal, , 

We developed a block-structured static adaptive mesh refinement (AMR) CFD code for the aerodynamics simulation using the lattice Boltzmann method on GPU supercomputers. The data structure of AMR was based on the forest-of-octrees, and the domain partitioning algorithm was based on space-filling curves (SFCs). To reduce the halo data communication, we introduced the tree cutting approach, which divided the global domains with a few octrees into small sub-domains with many octrees, leading to a hierarchical domain partitioning approach with the coarse structured block and the fine SFC partitioning in each block. The tree cutting improved the locality of the sub-divided domain, and reduced both the amount of communication data and the number of connections of the halo communication. In the strong scaling test on the Tesla V100 GPU supercomputer, the tree cutting approach showed 1.82 speedup at the performance of 2207 MLUPS (mega-lattice update per second) on 128 GPUs.