https://hgpu.org/?p=1519
Evaluation of parallel particle swarm optimization algorithms within the CUDA architecture