A GPGPU Physarum Cellular Automaton Model

Scientists have been gaining inspiration from several natural processes and systems to find fine solutions in many complex hard to solve engineering problems for many years now. Nevertheless, most of these natural systems suffer from great amount of time to perform; thus, scientists are seeking for computational tools and methods that could encapsulate in a conscious way nature’s genius, dealing at the same moment with time complexity. In this conquest, Cellular Automata (CA) proposed long time ago by John von Neumann, can be considered as a promising candidate. CA have the ability to capture the essential features of systems in which global complicated behavior emerges from the collective effect of simple components, which interact locally. These characteristics are immanent in many natural systems; namely Physarum polycephalum,an amoeba, is such a system. This simple organism presents the intelligence of finding effective solutions to demanding engineering problems such as shortest path(s) problems, various graph problems, evaluation of transport networks or even robotic control. In this paper, we move forward by taking advantage of a Graphical Processing Unit (GPU) and the Compute Unified Device Architecture (CUDA) programming model, to make use of the CA inherit parallelism when biomimicking the behavior of P. polycephalum in maze, providing the ability to find the minimum path between two spots. In this way we are able to produce a virtual easy-to-access lab speeding up significantly the biological paradigm when modeled by CA implemented in General Purpose computing on Graphics Processing Units (GPGPU) environment