Search

The Cellular Automata paradigm is an efficient tool to model and study complex systems such as traffic simulation, lava flows and swarm based behaviour. In addition, cellular automata can be profitably used in many mathematical problems such as differential equations and chaos theory. Due to their inherent parallel nature, cellular automata can be efficiently parallelized among a set of computing nodes in order to scale and speed up their execution. This paper presents a preliminary study on different parallelizzation techniques for structured grid models such as cellular automata on distributed memory architectures. In particular, three strategies are presented and compared in order to evaluate their efficiency in terms of speed-up. An experimental section shows the performance achieved by the three strategies when a real-life application, namely the SciddicaT cellular automata model for debris-flows simulation, is adopted.

The distribution of computational load over different processing elements represents an important issue in parallel computing [1]. This is particularly relevant in the case of parallel execution of structured grid computational models, such as Cellular Automata (CA) [2], where the domain space is partitioned in region assigned to the parallel computing nodes. Load balancing techniques are particularly effective when structured grid computational models are exploited to simulate topologically connected physical phenomena like lava or debris flows (e.g., [3]), in which the evolution develops in a usually small sub-region of the domain. In this work, we present a dynamic load balancing technique that can provide performance improvements in structured grid model development on distributed memory architectures by adopting the MPI technology. First tests have demonstrated the usefulness of the feature in appreciably reducing execution times in comparison with not-balanced parallel versions.