RAPIDS API Docs
Access our current docs for the RAPIDS projects below. Docs are available in both “stable” and “nightly” versions. The description of each is below to help select the docs that fit your needs.
- STABLE
- Current release docs; considered to be stable.
- NIGHTLY
- Work-in-progress release docs; considered to be unstable and released nightly.
- LEGACY
- Previous release docs; available for reference.
RAPIDS APIs
cuDF
cuDF is a Python GPU DataFrame library (built on the Apache Arrow columnar memory format) for loading, joining, aggregating, filtering, and otherwise manipulating data.
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
dask-cuDF
Dask-cuDF extends Dask where necessary to allow its DataFrame partitions to be processed using cuDF GPU DataFrames instead of Pandas DataFrames. Dask-cuDF is used to leverage multiple gpus and multiple nodes for more performance at larger scales
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
cuML
cuML is a suite of libraries that implement machine learning algorithms and mathematical primitives functions that share compatible APIs with other RAPIDS projects.
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
cuGraph
cuGraph is a GPU accelerated graph analytics library, with functionality like NetworkX, which is seamlessly integrated into the RAPIDS data science platform. cuGraph supports GNNs with PyG, DGL packages and cugraph-ops (for low level optimized operations) as well as cugraph-service for analytics on a remote graph and WHOLEGRAPH for memory management.
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
cuxfilter
cuxfilter acts as a connector library, which provides the connections between different visualization libraries and a GPU dataframe without much hassle. This also allows the user to use charts from different libraries in a single dashboard, while also providing the interaction.
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
cuSpatial
cuSpatial is a GPU-accelerated vector GIS library including binary predicates (DE-9IM), point-in-polygon, spatial join, distances, and trajectory analysis.
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
cuProj
cuProj is a GPU-accelerated geographic and geodetic coordinate transformation library which supports projecting coordinates between coordinate reference systems (CRSes), compatible with PyProj.
DOCS stable (23.12) | nightly (24.02)
LINKS changelog | github
cusignal
cuSignal functionality has been moved to CuPy. Please see the CuPy documentation for more information.
DOCS
LINKS changelog | github
Java + cuDF
Java bindings for the cuDF library.
DOCS stable (23.12) | legacy (23.10)
LINKS changelog | github
cuCIM
The RAPIDS cuCIM is an extensible toolkit designed to provide GPU accelerated I/O, computer vision & image processing primitives for N-Dimensional images with a focus on biomedical imaging.
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
KvikIO
KvikIO is a Python and C++ library for high performance file IO using GPUDirect Storage (GDS).
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
RAFT
RAFT contains fundamental widely-used algorithms and primitives for vector search, machine learning, and information retrieval.
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
Dask-CUDA
Various utilities to improve deployment and management of Dask workers on CUDA-enabled systems.
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
RMM
RAPIDS Memory Manager (RMM) is a central place for all device memory allocations in cuDF (C++ and Python) and other RAPIDS libraries. In addition, it is a replacement allocator for CUDA Device Memory (and CUDA Managed Memory) and a pool allocator to make CUDA device memory allocation / deallocation faster and asynchronous.
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
RAPIDS Libraries
libcudf
libcudf is a C/C++ CUDA library for implementing standard dataframe operations.
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
libcuspatial
libcuspatial is a GPU-accelerated header-only C++ vector GIS library including binary predicates (DE-9IM), point-in-polygon, spatial join, distances, and trajectory analysis.
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
libcuproj
libcuproj is a C++ header-only library for GPU-accelerated geographic and geodetic coordinate transformation library which supports projecting coordinates between coordinate reference systems (CRSes), similar to PROJ.
DOCS stable (23.12) | nightly (24.02)
LINKS changelog | github
libcuml
libcuml is a C/C++ CUDA library for cuML.
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
librmm
RAPIDS Memory Manager (RMM) is a central place for all device memory allocations in cuDF (C++ and Python) and other RAPIDS libraries. In addition, it is a replacement allocator for CUDA Device Memory (and CUDA Managed Memory) and a pool allocator to make CUDA device memory allocation / deallocation faster and asynchronous.
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
libkvikio
libkvikio is a C++ header-only library providing bindings to cuFile, which enables GPUDirect Storage (GDS).
DOCS stable (23.12) | nightly (24.02) | legacy (23.10)
LINKS changelog | github
libucxx
UCXX is an object-oriented C++ interface for UCX, with native support for Python bindings.
DOCS nightly (0.36)
LINKS github
rapids-cmake
This is a collection of CMake modules that are useful for all CUDA RAPIDS projects. By sharing the code in a single place it makes rolling out CMake fixes easier.