8282

Increasing the performance of AllToAll variant of self-organizing migration algorithm using CUDA

Michal Pavlech, Jan Seckar
Faculty of Applied Informatics, Tomas Bata University in Zlin, nam. T.G.Masaryka 5555, 760 01 Zlin, Czech Republic
Evolutionary Computing (EC ’12), 2012

@article{pavlech2012increasing,

   title={Increasing the performance of AllToAll variant of self-organizing migration algorithm using CUDA},

   author={Pavlech, Michal and Seckar, Jan},

   year={2012}

}

Download Download (PDF)   View View   Source Source   

552

views

Modern graphics processing units offer general purpose parallel computing capabilities. Thus they have become a relatively low cost alternative for applications requiring extensive parallel computations. Evolutionary algorithms are especially well suited for parallel SIMD architecture. This paper deals with the modification of AllToAll variation of self-organizing migration algorithm, which has high computational demand for one round of algorithm, using the CUDA framework. The main goal is to speedup performance of the algorithm in comparison to CPU implementations.
No votes yet.
Please wait...

* * *

* * *

HGPU group © 2010-2017 hgpu.org

All rights belong to the respective authors

Contact us: