Supercomputing pp 768-781 | Cite as

Finite element methods on parallel and vector computers application in fluid dynamics

  • J. Erhel
Session 9A: Algorithms, Architectures And Performance I
Part of the Lecture Notes in Computer Science book series (LNCS, volume 297)


The experiences described in this paper show that finite element methods can be vectorized efficiently, though they are slightly penalized by the need of indirect addressing. The extension to 3-D problems is currently under investigation. The main question is to find a graph coloring algorithm to get a small number of colors, with enough elements in each color. A useful heuristic is to freeze in advance the maximal size of each group, thus the vector length. Implicit schemes without storage of the matrix are also examined, to get a good tradeoff between memory requirements and CPU time.

Concerning the parallel algorithms, the subdomain approach seems more efficient. But the microtasking approach should give better results on larger problems, with a larger granularity. It is easier to implement and does not require a preliminary partitionning.

However, we plan to investigate partitionning methods, with more general configurations (variable number of neighbors), using general graph separation algorithms. Simulated annealing techniques are also considered to optimize locally the separator length.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    A. Dervieux, J.A. Desideri, F. Fezoui, B. Palmerio, J.P. Rosenblum Euler calculations by upwind finite element methods and adaptive mesh algorithms. GAMM Workshop, France 1986.Google Scholar
  2. [2]
    CRAY-XMP reference manual Cray Research, Inc.Google Scholar
  3. [3]
    CRAY-2 reference manual Cray Research, Inc.Google Scholar
  4. [4]
    ETA10 reference manual ETA Systems, Inc.Google Scholar
  5. [5]
    A. Jameson, T.J. Baker Improvements to the aircraft Euler method AIAA 25th aerospace sciences meeting, January 1987, USAGoogle Scholar
  6. [6]
    T. Hugues, R. Ferencz, J. Hallquist Large-scale vectorized implicit calculations in solid mechanics on a Cray X-MP/48 utilizing EBE preconditioned conjugate gradients Stanford University Report.Google Scholar
  7. [7]
    Cray X-MP microtasking user guide Cray Research, Inc.Google Scholar
  8. [8]
    Cray X-MP and CRAY-2 multitasking user guides Cray Research, Inc.Google Scholar
  9. [9]
    Proceedings of first international symposium on domain decomposition methods for partial differential equations, January 1987, FranceGoogle Scholar
  10. [10]
    G. Johnson, J. Swisshelm, D. Pryor, J. Ziebarth Multitasked embedded multigrid for three-dimensional flow simulation. 10th international conference on numerical methods in fluid mechanics, June 1986, China.Google Scholar
  11. [11]
    F. Fezoui, B. Stoufflet, J. Periaux, A. Dervieux Implicit high-order upwind finite element schemes for the Euler equations 4th International Symposium on numerical methods in engineering, March 1986, USAGoogle Scholar
  12. [12]
    F. Angrand, J. Erhel Simulation par éléments finis d'écoulements compressibles sur calculateurs vectoriels INRIA Research report no 622, February 1987.Google Scholar
  13. [13]
    IBM-VF reference manualGoogle Scholar
  14. [14]
    J. Erhel INRIA Research report, to appearGoogle Scholar
  15. [15]
    D. Kuck, R. Kuhn, D. Padua, B. Leasure, M. Wolfe Dependence graphs and compiler optimizations Proc. 8th ACM Symp. on Principles of Programming Languages, USA, January 1981.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1988

Authors and Affiliations

  • J. Erhel
    • 1
  1. 1.INRIA — Domaine de VoluceauLe Chesnay CédexFrance

Personalised recommendations