Abstract
The solution of finite element problems with irregular geometries on a parallel computer of the hypercube type (MIMD, distributed memory) is considered. The technique of scattering the decomposition is found to be easy to implement and to effectively load balance the computation.
Similar content being viewed by others
References
Carey, G. F., and Oden, J. (1983).Finite Elements, Vol. 1–6, Prentice Hall, Englewood Cliffs, New Jersey.
Dahlquist, G., and Björk, Å., (1974).Numerical Methods, Prentice Hall, Englewood Cliffs, New Jersey.
Fox, G. (1984). Proceedings of IEEE Compucon Conference, February 28.
Fox, G. (1985).IEEE Trans. N.P.S.S. 34.
Fox, G., and Otto, S. (1984).Phys. Today,May.
Fox, G.,et al. (1985). The Caltech Concurrent Computation Program, inProceedings of the 1985 ASME International Computers in Engineering Conference, August 4–8, Boston; published by ASME.
Hageman, A., and Young, M. (1983).Applied Iterative Methods, Academic, New York.
Johnson, M. (1986). The specification of CrOS III, C3P memo 253, February.
Lapidus, L., and Pinder, F. (1984).Numerical Solution of Partial Differential Equations in Science and Engineering, Wiley, New York.
Salmon, J. (1984). Binary grey codes and the mapping of a physical lattice into a hypercube, C3P memo 51, January.
Seitz, C. (1984). CACM, December.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Morison, R., Otto, S. The scattered decomposition for finite elements. J Sci Comput 2, 59–76 (1987). https://doi.org/10.1007/BF01061512
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF01061512