Summary
Loosely coupled multiprocessor systems seem to offer an interesting alternative to the solution of large numerical problems. It is in the context of such an investigation that we treat here a special parallel-processing case concerning the solution of a numerical problem on two independent processors including simultaneous input-output operations. We present a short discussion of the underlying numerical algorithm, a modelling approach to the parallel computing system and a comparison of the theoretically obtained results with simulation and experimental results. The experimental setting is the XANTHOS multi-microprocessor system implemented at the Laboratoire de Recherche en Informatique, Université de Paris-Sud.
Similar content being viewed by others
References
Cinlar, E.: Introduction to stochastic processes. J. Wiley 1977
Gelenbe, E.: Parallel computation of partial differential equations: a modelling approach. 19th IEEE Conference on Decision and Control, November 1980
Gelenbe, E., Geyl, J.D., Gibergues, O., Horlait, E.: Réseau local à diffusion de pacquets sur fibre optique: le système XANTHOS. Rapport de Recherche n∘ 63, L.R.I., Orsay, 1979
Gelenbe, E., Lichnewsky, A., Staphylopatis, A.: Experience with the parallel solution of partial differential equations on a distributed computer system. Rapport de Recherche n∘ 81, L.R.I., Orsay, 1980
Gelenbe, E., Mitrani, I. : Analysis and synthesis of computer systems. Academic Press 1980
Heller, D.: Parallel algorithms. SIAM Rev. 20, 740–777 (1978)
Kuck, D.J.: A survey of parallel machine organization and programming.
Kuhn, R.H.: Efficient mapping of algorithms to single stage interconnections. Department of Computer Science, University of Illinois, Urbana Champaign
Lichnewsky, A.: Sur la résolution de systèmes linéaires issus de la méthode des éléments finis par une machine ≪multiprocesseur≫. Département de Mathématique, Orsay, 1980
Richtmyer, R.D., Morton, K.W.: Difference methods for initial value problems. J. Wiley 1957
Author information
Authors and Affiliations
Additional information
This work was supported by a DGRST Research Scholarship at Université Paris-Sud
Rights and permissions
About this article
Cite this article
Staphylopatis, A. Performance considerations in the parallel execution of numerical algorithms on two processors. Acta Informatica 17, 311–325 (1982). https://doi.org/10.1007/BF00264356
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF00264356