[ARCHIVED] Cooperative primitives for CUDA C++. See https://github.com/NVIDIA/cccl
CUDA Kernel Benchmarking Library