Fast Numerical Simulation and Distributed Visualization of Gas-Liquid-Solid Multiphase Thermal Flow on GPU-CPU Cooperating Cluster

2017 Fiscal Year Final Research Report

• Project/Area Number: 15K00167
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: High performance computing
• Research Institution: University of Yamanashi
• Principal Investigator: ANDO Hidetoshi 山梨大学, 大学院総合研究部, 准教授 (50221742)
• Co-Investigator(Kenkyū-buntansha): 鳥山 孝司 山梨大学, 大学院総合研究部, 准教授 (50313789)
• Project Period (FY): 2015-04-01 – 2018-03-31
• Keywords: 高性能計算 / 可視化 / 多相流体シミュレーション / GPU

Outline of Final Research Achievements

We have developed a method where GPU and CPU effectively cooperate to achieve large-scale computation under small size GPU cluster, by using CPU side of memory from GPU. Proposed method realized fluid simulation using higher resolution grid while minimizing data communication delay between GPU and CPU. Experiments showed our implementation of multiphase thermal flow computation could handle various thermal flow simulations under many different kinds of situation. We have introduced the photon mapping method for advanced distributed visualization, and also for highly accurate thermal radiation computation.

Free Research Field

GPUコンピューティング