https://hgpu.org/?p=12599
Multi-Threaded Automatic Integration Using OpenMP and CUDA