New Ideas in Parallel Metaheuristics on GPU: Systolic Genetic Search

摘要: This chapter presents an in-depth study of a novel parallel optimization algorithm specially designed to run on Graphic Processing Units (GPUs). The underlying operation relates systolic computing and is inspired by the contraction heart that makes possible blood circulation. algorithm, called Systolic Genetic Search (SGS), based synchronous circulation solutions through grid processing units tries profit from architecture GPUs achieve high time performance. SGS has shown not only numerically outperform random search two genetic algorithms for solving Knapsack Problem over set increasingly sized instances, but also its implementation can obtain runtime reduction that, depending GPU technology used, reach more than 100 times. A performance four different been conducted show impact Nvidia’s compute capabilities runtimes algorithm.