cuda.coop: Cooperative Algorithms#

The cuda.coop library provides cooperative algorithms that operate at the level of blocks and warps. It is designed to be used within Numba CUDA kernels.

Here’s an example showing how to use the cuda.coop library to perform block-level reduction within a Numba CUDA kernel.

Example Collections#

For complete runnable examples and more advanced usage patterns, see our full collection of examples.

External API References#