Impact of asynchronism on GPU accelerated parallel iterative computations

Authors:
Sylvain Contassot-Vivier;Thomas Jost;Stéphane Vialle
Affiliations:
Loria, University Henri Poincaré, Nancy, France;AlGorille INRIA Project Team, France;AlGorille INRIA Project Team, France
Venue:
PARA'10 Proceedings of the 10th international conference on Applied Parallel and Scientific Computing - Volume Part I
Year:
2010

Citing 11
Cited 0

Optimal Asynchronous Newton Method for the Solution of Nonlinear Equations

Journal of the ACM (JACM)
Parallel and distributed computation: numerical methods

Parallel and distributed computation: numerical methods
BI-CGSTAB: a fast and smoothly converging variant of BI-CG for the solution of nonsymmetric linear systems

SIAM Journal on Scientific and Statistical Computing
On the theory and practice of multisplitting methods in parallel computation

Computing
Implicit-Explicit Parallel Asynchronous Solver of Parabolic PDEs

SIAM Journal on Scientific Computing
Convergence of partially asynchronous block quasi-Newton methods for nonlinear systems of equations

Journal of Computational and Applied Mathematics
Asynchronous Iterative Methods for Multiprocessors

Journal of the ACM (JACM)
On asynchronous iterations

Journal of Computational and Applied Mathematics - Special issue on numerical analysis 2000 Vol. III: linear algebra
Analysis of Asynchronous Polynomial Root Finding Methods on a Distributed Memory Multicomputer

IEEE Transactions on Parallel and Distributed Systems
Evaluation of the asynchronous iterative algorithms in the context of distant heterogeneous clusters

Parallel Computing
An Efficient and Robust Decentralized Algorithm for Detecting the Global Convergence in Asynchronous Iterative Algorithms

High Performance Computing for Computational Science - VECPAR 2008

Quantified Score

Hi-index	0.00

Visualization

Abstract

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).