Evolving FPGA based cellular automata

Cellular Automata architectures are attractive due to their fine grain parallelism, simple computational structures and local routing resources. Some researchers have used genetic algorithms to find CA that perform useful computations. The inherently parallel cellular automata model as well as the genetic algorithm are poorly suited to implementation on general purpose microprocessor based systems. Field Programmable Gate Arrays are an alternative that can provide significant speedup. This paper describes the Xilinx XC6216 Field Programmable Gate Array and how it is used to efficiently search a hybrid 2-state, 5-neighbour cellular automata rule space that exhibits computation universality. Its application to an image processing application, binary texture analysis, is discussed.