RAPIDS
Accelerate common data science workloads on GPUs
RAPIDS is a collection of libraries that enable you to take advantage of NVIDIA GPUs to accelerate machine learning workflows. Minimal changes are required to transition from familiar pandas and scikit-learn to GPU accelerated code! For more information on RAPIDS, see “Getting Started” in the RAPIDS docs.
Use RAPIDS on a Single GPU
Use RAPIDS to do machine learning on a single GPU
Use RAPIDS on a GPU Cluster
Use RAPIDS to do machine learning on a GPU cluster