Research Article

Pipelines on heterogeneous systems: models and tools

  1. F. Almeida*,
  2. D. Gonzalez,
  3. L. M. Moreno,
  4. C. Rodriguez

Article first published online: 9 MAR 2005

DOI: 10.1002/cpe.886

Issue

Concurrency and Computation: Practice and Experience

Concurrency and Computation: Practice and Experience

Volume 17, Issue 9, pages 1173–1195, 10 August 2005

Additional Information

How to Cite

Almeida, F., Gonzalez, D., Moreno, L. M. and Rodriguez, C. (2005), Pipelines on heterogeneous systems: models and tools. Concurrency and Computation: Practice and Experience, 17: 1173–1195. doi: 10.1002/cpe.886

Author Information

  1. Dpto. Estadística, I. O. y Computación, Edificio Físicas/Matemáticas, Universidad de La Laguna, La Laguna, Tenerife, Spain

*Dpto. Estadística, I. O. y Computación, Edificio Físicas/Matemáticas, Universidad de La Laguna, La Laguna, Tenerife, Spain

Publication History

  1. Issue published online: 9 MAR 2005
  2. Article first published online: 9 MAR 2005
  3. Manuscript Accepted: 9 MAR 2004
  4. Manuscript Revised: 1 MAR 2004
  5. Manuscript Received: 29 MAY 2003

Funded by

  • EC (FEDER)
  • Spanish MCyT. Grant Numbers: TIC2002-04498-C05-05, TIC2002-04400-C03-03

SEARCH

SEARCH BY CITATION

ARTICLE TOOLS

Get PDF (327K)

Keywords:

  • heterogeneous systems;
  • pipelines;
  • analytical models

Abstract

We study the performance of pipeline algorithms in heterogeneous networks. The concept of heterogeneity is not only restricted to the differences in computational power of the nodes, but also refers to the network capabilities. We develop a skeleton tool that allows us an efficient block-cyclic mapping of pipelines on heterogeneous systems. The tool supports pipelines with a number of stages much larger than the number of physical processors available. We derive an analytical formula that allows us to predict the performance of pipelines in heterogeneous systems. According to the analytical complexity formula, numerical strategies to solve the optimal mapping problem are proposed. The computational results prove the accuracy of the predictions and effectiveness of the approach. Copyright © 2005 John Wiley & Sons, Ltd.

Get PDF (327K)