Nvidia's Hyper-Q: 2.5x speed up without rewriting your legacy CUDA applicationsblogs.nvidia.com1 point·rplnt··0 commentsOpen articleSaveView on HN