Abstract
New approach for solving of compressible fluid dynamic problems with complex geometry on Cartesian grids is proposed. It leads to algorithmic uniformity for whole domain and structured memory accesses which are essential for effective implementations on massively-parallel architectures – GPUs. Methods used are based on implicit scheme and LU-SGS method. Novel parallel algorithm for last one is proposed. In-depth analysis of CUDA+MPI implementation (interoperability issues, libraries tuning) scalable up to hundreds GPUs is performed.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
Jameson, A., Turkel, E.: Implicit schemes and LU decomposition. Math.of Comp 37, 385–397 (1981)
Lin, F., Zhenghong, G., Kan, H., Fang, X.: A multi-block viscous flow solver based on GPU parallel methodology. Comput. Fluids 95, 19–39 (2014)
Wright, M.J., Candler, G.V.: A data-parallel LU rexation method for reacting viscous flows. In: Ecer, A., Satofuka, N., Periaux, J., Taylor, S. (eds.) Parallel Computational Fluid Dynamics 1995, pp. 67–74. North-Holland, Amsterdam (1996)
Menshov, I., Pavlukhin, P.: Numerical Solution of Gas Dynamics Problems on Cartesian Grids with the Use of Hybrid Computing Systems. Preprint of KIAM RAS. vol. 92 (2014)
Menshov, I., Nakamura, Y.: On implicit godunov’s method with exactly linearized numerical flux. Comput. Fluids 29(6), 595–616 (2000)
Menshov, I., Nakamura, Y.: Hybrid explicit-implicit, unconditionally stable scheme for unsteady compressible flows. AIAA J. 42(3), 551–559 (2004)
Pavlukhin, P.: Parallel LU-SGS numerical method implementation for gas dynamics problems on GPU-accelerated computer systems. Vestn. of Lobachevsky State Univ. Nizhni novgorod 1, 213–218 (2013)
“K-100" Supercomputer. http://www.kiam.ru/MVS/resourses/k100.html
CUDA C/C++ Streams and Concurrency. http://on-demand.gputechconf.com/gtc-express/2011/presentations/StreamsAndConcurrencyWebinar.pdf
CUDA Toolkit Documentation. http://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__DEVICE.html
Lomonosov Supercomputer. http://hpc.msu.ru/?q=node/59
Acknowledgments
This work was supported by grant 14–01–31480 from the Russian Foundation for Basic Research.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2015 Springer International Publishing Switzerland
About this paper
Cite this paper
Pavlukhin, P., Menshov, I. (2015). On Implementation High-Scalable CFD Solvers for Hybrid Clusters with Massively-Parallel Architectures. In: Malyshkin, V. (eds) Parallel Computing Technologies. PaCT 2015. Lecture Notes in Computer Science(), vol 9251. Springer, Cham. https://doi.org/10.1007/978-3-319-21909-7_42
Download citation
DOI: https://doi.org/10.1007/978-3-319-21909-7_42
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-21908-0
Online ISBN: 978-3-319-21909-7
eBook Packages: Computer ScienceComputer Science (R0)