[ARCHIVED] Cooperative primitives for CUDA C++. See https://github.com/NVIDIA/cccl
Here you can share your experience with the project you are suggesting or its comparison with cub. Optional.
A valid email to send you a verification link when necessary or log in.
cub