Many problems in computer simulation of systems in science and engineering present potential for parallel implementations through one of the three major paradigms of algorithmic parallelism, geometric parallelism and processor farming. Static process scheduling techniques have been used successfully to exploit geometric and algorithmic parallelism, while dynamic process scheduling is better suited to dealing with the independent processes inherent in the process farming paradigm. This paper considers the application of parallel or multi-computers to a class of problems exhibiting spatial data dependency characteristic of the geometric paradigm. However, by using processor farming paradigm in conjunction with geometric decomposition, a dynamic scheduling technique is developed to suit the MIMD structure of the multi-computers. The specific problem chosen for the investigation of scheduling techniques is the computer simulation of Cellular Automaton models. A cellular automaton simulation, with artificially increased compute load per cell (in the form of number of simulated multiplies) is considered for parallelization. Such a simulation is representative of a class of recursive algorithms with local spatial dependency and fine granularity that may be encountered in biological applications, finite elements, certain problems in image analysis and computational geometry (2)- (5). These types of applications exhibit geometric parallelism and may be considered best suited to static scheduling. However, using dynamic scheduling, the MIMD structure of multicomputer networks is exploited, and comparison of both the schemes is given in the form of total timings and speedup. II. THE C.A. MODEL
[1]
Alexander Walsch.
Architecture and prototype of a real-time processor farm running at 1 MHz
,
2002
.
[2]
C. Moore,et al.
CELLULAR AUTOMATA: A DISCRETE UNIVERSE By ANDREW ILACHINSKI: 808 pp., £76.00, ISBN 981-02-4623-4 (World Scientific, Singapore, 2001).
,
2003
.
[3]
Mehrdad Ghaemi,et al.
A Novel Method for Simulating Cancer Growth
,
2010,
ACRI.
[4]
Samuel T. Chanson,et al.
Performance Models for the Processor Farm Paradigm
,
1997,
IEEE Trans. Parallel Distributed Syst..
[5]
Stephen A. Billings,et al.
Identification of Excitable Media Using Cellular Automata Models
,
2007,
Int. J. Bifurc. Chaos.
[6]
Emanuela Merelli,et al.
Bone Remodelling: A Complex Automata-Based Model Running in BioShape
,
2010,
ACRI.
[7]
Thomas L. Casavant,et al.
A Taxonomy of Scheduling in General-Purpose Distributed Computing Systems
,
1988,
IEEE Trans. Software Eng..
[8]
J PritchardD.
Transputer Applications on Supernode
,
1990
.
[9]
Maria Vittoria Avolio,et al.
Development and Calibration of a Preliminary Cellular Automata Model for Snow Avalanches
,
2010,
ACRI.
[10]
Mohammad Shakeel Laghari,et al.
Comparison of scheduling techniques for the parallel implementation of the Hough transform
,
1991,
Algorithms and Parallel VLSI Architectures.
[11]
Yoon Seok Yang,et al.
Parallel and Pipeline Processing for Block Cipher Algorithms on a Network-on-Chip
,
2009,
2009 Sixth International Conference on Information Technology: New Generations.