Parallelisation of Shallow Water Simulation for Heterogeneous Architectures

Michail Emmanouil Pappas
School of Computer Science, University of Manchester
University of Manchester, 2012


   title={Parallelisation of Shallow Water Simulation for Heterogeneous Architectures},

   author={Pappas, M.E.},



Download Download (PDF)   View View   Source Source   



This work presents the parallelisation of a shallow water simulation model. Two parallel implementations are developed. One is for a multi-core NUMA architecture, developed in OpenMP. The other one is for a many-core GPU-accelerated architecture and is developed in OpenCL. The parallelisation process is based on an iterative approach, starting off from a naive implementation. Each iteration involves the identification, analysis and improvement of a particular overhead. The process repeats until no further optimisations can be performed. The evaluation is based on the comparison of the two implementations. The final results show strengths and weaknesses in both implementations, as they are reflected from the differences of the two architectures.
No votes yet.
Please wait...

* * *

* * *

HGPU group © 2010-2021 hgpu.org

All rights belong to the respective authors

Contact us: