A review of genetic algorithms and parallel genetic algorithms on Graphics Processing Unit (GPU)

Genetic Algorithms (GAs) is proven to be effective in solving many optimization tasks. GAs is one of the optimization tools used widely in solving problems based on natural selection and genetics. This paper is intended to cover the study of GA and parallel GA and analyses its usage in CPU and GPU. One of the popular ways to speed up the processing time was by running them as parallel. The idea of parallel GAs may refer to an algorithm that works by dividing large problem into smaller tasks. Broad literature review in this paper includes a categorization of the GA operations that involved with some theories and techniques used in GA, presented with the aid of diagrams. This review attempts to study and analyse the behaviour of GA and parallel GA categories to work in GPU depending on the type of genetic algorithm. Parallel GA for GPU covers the architecture of Compute Unified Device Architecture (CUDA)