GPU implementation of physarum cellular automata model

Abstract

In the past few decades, there is an increasing number of publications which show that solutions to complex mathematical problems can be found by applying unconventional computing methods. Among other examples, the plasmodium of Physarum Polycephalum has been intensively used for solving shortest path(s) problem, various graph problems, evaluation of transport networks, robotic control and many other engineering applications. In this paper we coupled the computing abilities of slime mould with one of the most powerful parallel computational models, namely Cellular Automata (CAs). CAs can capture the essential features of systems which global behavior emerges from the collective effect of simple components, which interact locally. Moreover, a Graphical Processing Unit (GPU) implementation will exploit the prominent feature of parallelism that CA structures inherently possess in contrast to the serial computers, thus accelerating the response of the proposed model. As a result, a slime mould CA based model in graphics processing unit (GPU) using CUDA programming model to describe and mimic the behavior of a plasmodium in a maze. In this way we are able to produce a virtual lab speeding up significantly the biological paradigm in GPU.