Abstract
As the number of parallel processors in massively parallel systems increases, the need for concurrent tasks or processes also increases. In this study the author investigates the total parallelism inherent in NWP algorithms. The technique employed is to implement a well known, highly simplified algorithm for solving the shallow water equations in one of the languages of dataflow systems, SISAL. In this form all extent parallelism can be detected by the compiler and made accessible to the dataflow computational environment. In this case the environment is a dataflow simulator which produces detailed diagnostics about the parallelism in the algorithm. These results are then compared to similar values obtained from a conventional SIMD system, a Cray X/MP.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Geerd-R. Hoffmann, P. N. Swarztrauber, & R. A. Sweet, “Aspects of Using Multiprocessors for Meteorological Modelling,” in Multiprocessing in Meteorological Models Hoffmann, Geerd-R. and D. F. Snelling eds. (Springer-Verlag, Berlin: 1988).
R. Sadourney, “The Dynamics of Finite Difference Models of the Shallow Water Equations,” Journal of Atmospheric Sciences No. 32 (1975) pp.680–689
Snelling, D. F. and D. A. Tanqueray, “Performance Measurements of the Shallow Water Equations on the FPS T Series,” submitted to CONPAR 88.
J. Gurd, “Dataflow Architectures and Implicit Parallel Programming,” in Multiprocessing in Meteorological Models Hoffmann, Geerd-R. and D. F. Snelling eds. (Springer-Verlag, Berlin: 1988).
J. Gurd et al., “The Manchester Prototype Dataflow Computer,” in Communications of the ACM, 28: 1, (1985) pp. 34–52.
Special issue on Dataflow computing, IEEE Computer 15:2, (1982).
J. R. McGraw et al. SISAL — Streams and Iteration in a Single Assignment Language Language Reference Manual V1.2 Lawrence Livermore National Laboratory, (1983)
J. Sargent, Simulator Users Guide Dataflow Research Group, Department of Computer Science, University of Manchester (1986)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1990 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Snelling, D.F. (1990). Inherent Parallelism in Numerical Weather Prediction Algorithms. In: Hoffmann, GR., Maretis, D.K. (eds) The Dawn of Massively Parallel Processing in Meteorology. Topics in Atmospheric and Oceanic Sciences. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-84020-3_11
Download citation
DOI: https://doi.org/10.1007/978-3-642-84020-3_11
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-84022-7
Online ISBN: 978-3-642-84020-3
eBook Packages: Springer Book Archive