Collection of multi-GPU, distributed memory algorithms. RapidsMPF provides a unified framework for asynchronous, multi-GPU pipelines using simple streaming primitives built on RAPIDS components.
- Getting Started (nightly)
- Background (nightly)
- Configuration Options (nightly)
- Python API Reference (nightly)
- C++ API Reference (nightly)
- Glossary (nightly)
git clone https://github.com/rapidsai/rapidsmpf.git
cd rapidsmpf
mamba env create --name rapidsmpf-dev --file conda/environments/all_cuda-131_arch-$(uname -m).yaml
./build.shSee the Getting Started guide (nightly) for debug builds, AddressSanitizer, MPI/UCX test suites, and rrun launcher details.