High performance computing for global optimization problems

Abstract

In the present work, the multiextremal optimization problems and a high-performance parallel algorithm for solving these ones are considered. The investigation of the algorithm scalability has been carried out on the problem class, in which the computation costs of the functions depended on the iteration point. The algorithm proposed in the present work can utilize the CPUs (for solving more complex subproblems) as well as the GPUs (for solving the simple subproblems). The results of numerical experiments demonstrating the speedup when solving a series of multiextremal constrained problems are presented.