Performance Analysis for Parallel Adaptive FEM on SMP Clusters

  • Judith Hippold
  • Gudula Rünger
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3732)

Abstract

The parallel implementation of 3-dimensional adaptive finite element methods (FEM) on recent clusters of SMPs has to deal with dynamic load increase due to adaptive refinement which might result in load imbalances. Additionally, the irregular behavior of adaptive FEM causes further impacts like varying cache effects or communication costs which make a redistribution strategy more complex. In this paper we investigate the interaction of these effects and the performance behavior of parallel adaptive FEM and present several performance results on recent SMP clusters.

Keywords

Execution Time Cluster Node Average Execution Time Load Imbalance Refinement Level 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
This is a preview of subscription content, log in to check access.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Beuchler, S., Meyer, A.: SPC-PM3AdH v1.0, Programmer’s Manual. Technical Report SFB393/01-08. Chemnitz University of Technology (2001)Google Scholar
  2. 2.
    Hippold, J., Meyer, A., Rünger, G.: An Adaptive, 3-Dimensional, Hexahedral Finite Element Implementation for Distributed Memory. In: Bubak, M., van Albada, G.D., Sloot, P.M.A., Dongarra, J. (eds.) ICCS 2004. LNCS, vol. 3037, pp. 146–154. Springer, Heidelberg (2004)CrossRefGoogle Scholar
  3. 3.
    Hippold, J., Rünger, G.: A Data Management and Communication Layer for Adaptive, Hexahedral FEM. In: Danelutto, M., Vanneschi, M., Laforenza, D. (eds.) Euro-Par 2004. LNCS, vol. 3149, pp. 718–725. Springer, Heidelberg (2004)CrossRefGoogle Scholar
  4. 4.
  5. 5.
    Huse, L.P., Omanga, K., Bugge, H., Ry, H., Haugsdal, A.T., Rustad, E.: ScaMPI - Design and ImplementationGoogle Scholar
  6. 6.
    Meyer, A.: A parallel preconditioned conjugate gradient method using domain decomposition and inexact solvers on each subdomain. Computing 45, 217–234 (1990)MATHCrossRefMathSciNetGoogle Scholar
  7. 7.
    Meyer, A.: Parallel Large Scale Finite Element Computations. In: Cooperman, G., Michler, G., Vinck, H. (eds.) ICALP 1986. LNCIS, vol. 226, pp. 91–100. Springer, Heidelberg (1997)Google Scholar
  8. 8.
    Meyer, A., Michael, D.: A modern approach to the solution of problems of classic elasto– plasticity on parallel computers. Numerical Linear Algebra and Application 4(3), 205–221 (1997)MATHCrossRefMathSciNetGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Judith Hippold
    • 1
  • Gudula Rünger
    • 1
  1. 1.Department of Computer ScienceChemnitz University of TechnologyChemnitzGermany

Personalised recommendations