Simulating and benchmarking the shallow-water fluid dynamical equations on multiple graphical processing units

The shallow-water model equations provide a simple yet realistic benchmark problem in computational fluid dynamics (CFD) that can be implemented on a variety of computational platforms. Graphical Processing Units can be used to accelerate such problems either singly using a data parallel decompositional scheme or with multiple devices using a domain decompositional approach. We implement the SW equations on a range of modern GPUs with both parallel schemes and report on the typical performance. We compare integer optimised GPUs and very modern floating-point intensive GPU devices such as NVidia's Kepler K20X, and also investigate different m-GPU communication methods for geometric decompositions. We give detailed performance results and a summary of the main parallelisation issues.

[1]  Tuomo Kauranne,et al.  On the Parallelization of Global Spectral Weather Models , 1994, Parallel Comput..

[2]  K. Hawick,et al.  Visual Simulation of a Multi-Species Coloured Lattice Gas Model , 2012 .

[3]  Martin Berzins,et al.  Large Scale Parallel Solution of Incompressible Flow Problems Using Uintah and Hypre , 2013, 2013 13th IEEE/ACM International Symposium on Cluster, Cloud, and Grid Computing.

[4]  Xiaolei Zou,et al.  Impact of GMS-5 and GOES-9 Satellite-Derived Winds on the Prediction of a NORPEX Extratropical Cyclone , 2002 .

[5]  Kenneth A � Hawick,et al.  Parallelisation of the Uni � ed Model Data Assimilation Scheme , 2009 .

[6]  Francis Sullivan.  Turing’s Cathedral: The Origins of the Digital Universe , 2012 .

[7]  C. Fletcher.  Computational techniques for fluid dynamics , 1992 .

[8]  Kenneth A. Hawick,et al.  Hypercubic storage layout and transforms in arbitrary dimensions using GPUs and CUDA , 2011, Concurr. Comput. Pract. Exp..

[9]  Clive A. J. Fletcher,et al.  Computational Techniques for Fluid Dynamics: Specific Techniques for Different Flow Categories , 1990 .

[10]  Michael Griebel,et al.  Solving incompressible two-phase flows on multi-GPU clusters , 2013 .

[11]  J. Peixoto,et al.  Physics of climate , 1984 .

[12]  Kenneth A. Hawick,et al.  Comparison of GPU architectures for asynchronous communication with finite‐differencing applications , 2012, Concurr. Comput. Pract. Exp..

[13]  Paulius Micikevicius,et al.  3D finite difference computation on GPUs using CUDA , 2009, GPGPU-2.

[14]  K. A. Hawick,et al.  Visualising Multi-Phase Lattice Gas Fluid Layering Simulations , 2011 .

[15]  David H. Bailey,et al.  The Nas Parallel Benchmarks , 1991, Int. J. High Perform. Comput. Appl..

[16]  Manuel Jesús Castro Díaz,et al.  A multi‐GPU shallow‐water simulation with transport of contaminants , 2013, Concurr. Comput. Pract. Exp..

[17]  Gregory A. Kiker,et al.  An object-oriented hydrologic model for humid, shallow water-table environments , 2008 .

[18]  Michael Griebel,et al.  Numerical Simulation in Fluid Dynamics: A Practical Introduction , 1997 .

[19]  Ken A. Hawick,et al.  Performance Tradeoff Spectrum of Integer and Floating Point Applications , 2013 .

[20]  Doug L. James,et al.  Real time physics: class notes , 2008, SIGGRAPH '08.

[21]  C. Bailey.  Mathematical modelling of shallow water flows with application to Moreton Bay, Brisbane , 2010 .

[22]  G. Meehl,et al.  A Unified Modeling Approach to Climate System Prediction , 2009 .

[23]  Kenneth A. Hawick,et al.  Data-Parallelism and GPUs for Lattice Gas Fluid Simulations , 2010, PDPTA.

[24]  Kenneth A. Hawick,et al.  Exploiting graphical processing units for data‐parallel scientific applications , 2009, Concurr. Comput. Pract. Exp..

[25]  Luca Faust,et al.  Physical Fluid Dynamics , 2016 .

[26]  W. Mccomb,et al.  The physics of fluid turbulence. , 1990 .