Explicit integration with GPU acceleration for large kinetic networks

JOURNAL OF COMPUTATIONAL PHYSICS

卷:302

Article

Brock, Benjamin^1,3  Belt, Andrew^2,3  Billings, Jay Jay^3,5  Guidry, Mike^2,3,4 

[1] Univ Tennessee, Dept Elect Engn & Comp Sci, Knoxville, TN 37996 USA

[2] Univ Tennessee, Dept Phys & Astron, Knoxville, TN 37996 USA

[3] Oak Ridge Natl Lab, Comp Sci & Math Div, Oak Ridge, TN 37830 USA

[4] Oak Ridge Natl Lab, Div Phys, Oak Ridge, TN 37830 USA

[5] Univ Tennessee, Bredesen Ctr Interdisciplinary Res & Grad Educ, Knoxville, TN 37996 USA

关键词: Ordinary differential equations;    Reaction networks;    Stiffness;    Reactive flows;    Nucleosynthesis;    Combustion;   

DOI  :  10.1016/j.jcp.2015.09.013

来源: Elsevier

【 摘 要 】

We demonstrate the first implementation of recently-developed fast explicit kinetic integration algorithms on modern graphics processing unit (GPU) accelerators. Taking as a generic test case a Type Ia supernova explosion with an extremely stiff thermonuclear network having 150 isotopic species and 1604 reactions coupled to hydrodynamics using operator splitting, we demonstrate the capability to solve of order 100 realistic kinetic networks in parallel in the same time that standard implicit methods can solve a single such network on a CPU. This orders-of-magnitude decrease in computation time for solving systems of realistic kinetic networks implies that important coupled, multiphysics problems in various scientific and technical fields that were intractable, or could be simulated only with highly schematic kinetic networks, are now computationally feasible. (C) 2015 Elsevier Inc. All rights reserved.

【 授权许可】

Free