https://hgpu.org/?p=3536
Accelerating global sequence alignment using CUDA compatible multi-core GPU