Abstract

A parallel implementation of Genetic Programming is described, using the Bulk Synchronous Parallel Programming (BSP) model, as implemented by the Oxford BSP library. It is shown that considerable speedup of the GP execution can be achieved. As the complexity and the size of the problem increases, the actual speedup can be improved (assuming a constant number of processors), since the communication overhead is small compared with the parallel GP parts. The ease of use of BSP and the speedup achieved with the corresponding parallel implementation, suggests that GP researchers should consider BSP parallel implementations when dealing with time consuming problems.

Citations