Impact of asynchronism on GPU accelerated parallel iterative computations
Loria, University Henri Poincare, Nancy, France
Applied Parallel and Scientific Computing, Lecture Notes in Computer Science, Volume 7133/2012, 43-53, 2012
@article{contassot2012impact,
title={Impact of asynchronism on GPU accelerated parallel iterative computations},
author={Contassot-Vivier, S. and Jost, T. and Vialle, S.},
journal={Applied Parallel and Scientific Computing},
pages={43–53},
year={2012},
publisher={Springer}
}
We study the impact of asynchronism on parallel iterative algorithms in the particular context of local clusters of workstations including GPUs. The application test is a classical PDE problem of advection-diffusion-reaction in 3D. We propose an asynchronous version of a previously developed PDE solver using GPUs for the inner computations. The algorithm is tested with two kinds of clusters, a homogeneous one and a heterogeneous one (with different CPUs and GPUs).
March 22, 2012 by hgpu
