Parallelization efficiency versus stochasticity in simulation reaction–diffusion by cellular automata

Abstract

Due to a growing interest in chemical and biological phenomena, simulation of reaction-diffusion processes on micro level becomes urgently wanted. Asynchronous cellular automata (ACA) are promising mathematical models to be used as a base for creating computer simulation programs, which gives reason for investigation of the models capability. In particular, micro-level simulation requires to deal with very large ACA size. So, parallel implementation is inevitable, and, hence, achieving good parallelization efficiency is essential. Since parallelization efficiency depends on stochasticity (the degree of randomness) of the process under simulation, it is important to investigate their relations in order to create methods of developing ACA models with proper stochasticity values. In the paper the interrelation between stochasticity and parallelization efficiency is studied in the context of reaction-diffusion processes simulation on supercomputer with distributed memory. The results are illustrated by simulation a Large-scale process of wave front propagation.

Cellular Automata Modeling of Three-Dimensional Chitosan-Based Aerogels Fiberous Structures with Bezier Curves

“…Thus, cellular automata describes the behavior of large and complex systems using simple transition rules. In addition, CA-based models can be easily implemented using high-performance parallel calculations, which further increases the computational efficiency [ 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 ].…”