Parallel Implementation of PSO Algorithm Using GPGPU
Jaspreet Kaur, Satvir Singh, Sarabjeet Singh · 2016
The goal of this paper is to show how swarm intelligence inspired optimization algorithms can take benefit of the parallel computing mechanism supported by general purpose computing ability of a Graphical Processing Unit (GPU). In this paper, Particle Swarm Optimization (PSO) algorithm is implemented both in C (serial) and C-CUDA (parallel) and their performances are compared on a testbed of well-known optimization test functions. Simulation results showed that parallel implementation of PSO using C-CUDA searches near optimal solution in lesser time as compared to that of serial algorithm implemented using C.