Computing Strongly Connected Components in Parallel on CUDA