Applying parallel genetic algorithm to sorting problem

Researchers attempt to design the advanced computer architecture needed to develop new GA techniques that fully use the parallel capabilities of such powerful machine. The combination of GAs and massively parallel computing will combine population based search models with vast computational resources, which has the possibility of removing the computational bottleneck that prevents many GA systems from applying real-world problems in real time. In this paper, on the multiprocessor system FIN two kinds of parallel implementations of GAs, that is, fine-grained parallel GA and distributed GA are described and compared. Finally it is shown that a sorting problem can be solved using fine-grained parallel GAs.