https://hgpu.org/?p=24421
Performance Analysis and Improvement of Parallel Differential Evolution