Adaptive Kinetic-Fluid Solvers for Heterogeneous Computing Architectures
Dorodnicyn Computing Center of Russian Academy of Sciences, Vavilova-40, Moscow, 119333, Russia
arXiv:1503.00707 [physics.comp-ph], (2 Mar 2015)
@article{zabelok2015adaptive,
title={Adaptive Kinetic-Fluid Solvers for Heterogeneous Computing Architectures},
author={Zabelok, Sergey and Arslanbekov, Robert and Kolobov, Vladimir},
year={2015},
month={mar},
archivePrefix={"arXiv"},
primaryClass={physics.comp-ph}
}
This paper describes recent progress towards porting a Unified Flow Solver (UFS) to heterogeneous parallel computing. UFS is an adaptive kinetic-fluid simulation tool, which combines Adaptive Mesh Refinement (AMR) with automatic cell-by-cell selection of kinetic or fluid solvers based on continuum breakdown criteria. The main challenge of porting UFS to graphics processing units (GPUs) comes from the dynamically adapted mesh, which causes irregular data access. We describe the implementation of CUDA kernels for three modules in UFS: the direct Boltzmann solver using discrete velocity method (DVM), the Direct Simulation Monte Carlo (DSMC) module, and the Lattice Boltzmann Method (LBM) solver, all using octree Cartesian mesh with AMR. Double digit speedups on single GPU and good scaling for multi-GPU have been demonstrated.
March 3, 2015 by hgpu
