https://hgpu.org/?p=14299
Comparison between GPU and parallel CPU optimizations in viewshed analysis