GPU accelerated three dimensional unstructured geometric multigrid solver

Sebastian, J and Sivadasan, N and Banerjee, Raja (2014) GPU accelerated three dimensional unstructured geometric multigrid solver. In: International Conference on High Performance Computing and Simulation, HPCS 2014, 21-25 July, 2014, Bologna; Italy.

Full text not available from this repository. (Request a copy)

Abstract

Graphics processor units (GPUs) have started becoming an integral part of high performance computing. We develop a GPU based 3D-unstructured geometric multigrid solver, which is extensively used in Computational Fluid Dynamics (CFD) applications. Parallelization for GPUs is not straightforward because of the irregularity of the mesh. Using combination of graph coloring and greedy maximal independent set computations, we obtain significant performance improvements in the multigrid solver and its parallelization. We use NVIDIAs CUDA programming model for the implementation. In our experiments, we solve heat conduction problems on unstructured 3D meshes. Different schemes for implementing the multigrid algorithm are evaluated. For a mesh of size 1.6 million, our multigrid GPU implementation gives 24 times speed up compared to multigrid serial implementation and 1630 times speed up compared to non-multigrid serial implementation.

[error in script]
IITH Creators:
IITH CreatorsORCiD
Sivadasan, NUNSPECIFIED
Banerjee, Rajahttp://orcid.org/0000-0002-7163-1470
Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: Computational Fluid Dynamics; GPU Computing; High Performance Computing; Multigrid Flow Solver
Subjects: Physics > Mechanical and aerospace
Computer science > Big Data Analytics
Divisions: Department of Computer Science & Engineering
Department of Mechanical & Aerospace Engineering
Depositing User: Team Library
Date Deposited: 18 Nov 2014 09:07
Last Modified: 31 Aug 2017 03:53
URI: http://raiithold.iith.ac.in/id/eprint/854
Publisher URL: https://doi.org/10.1109/HPCSim.2014.6903663
Related URLs:

Actions (login required)

View Item View Item
Statistics for RAIITH ePrint 854 Statistics for this ePrint Item