Parallel Implementation of Vortex Element Method on CPUs and GPUs

The implementations of 2D vortex element method adapted to different types of parallel computers are considered. The developed MPI-implementation provides close to linear acceleration for small number of computational cores and approximately 40-times acceleration for 80-cores cluster when solving mo...

Full description

Saved in:
Bibliographic Details
Published in:Procedia computer science 2015, Vol.66, p.73-82
Main Authors: Kuzmina, Kseniia, Marchevsky, Ilia, Moreva, Victoriya
Format: Article
Language:English
Subjects:
fast multipole method
MPI
NVidia CUDA
parallel implementation
Vortex element
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The implementations of 2D vortex element method adapted to different types of parallel computers are considered. The developed MPI-implementation provides close to linear acceleration for small number of computational cores and approximately 40-times acceleration for 80-cores cluster when solving model problem. OpenMP-based modification allows to obtain 5% additional acceleration due to shared memory usage. Approximate fast multipole method usage reduces time of computations significantly: 11 times for the testmodel problem in sequential mode and 3.5 times in parallel mode for 16-cores cluster. The most efficient implementation of vortex element method is developed for GPUs using NVidia CUDA technology. Time of the model problem solving using single GeForce GTX 970 or Tesla C2070 accelerator is comparable with time of its solving on cluster when involving 30–40 cores of Intel Xeon E5450 CPUs.
ISSN:1877-0509
1877-0509
DOI:10.1016/j.procs.2015.11.010