RAPIDS in one line: RAPIDS executes end-to-end data pipelines entirely on GPUs.
RAPIDS is a set of libraries and APIs that runs on NVIDIA CUDA® GPUs. It offers GPU-enabled substitutes for popular Python libraries, such as cuDF for Pandas, cuML for scikit-learn, and cuGraph for NetworkX. These libraries use very similar APIs to the non-RAPIDS versions, so there’s less of a learning curve when moving code onto GPUs with RAPIDS.
RAPIDS aims to allow developers to run their entire data pipeline on GPUS, including doing ETL, visualization, model training, and inference. It’s fast because it relies not only on GPUs, but also on Numba and Apache Arrow.
Tools to support GPU-enabled machine learning training have been around for a while, but none of them enabled GPU-use over an entire machine learning pipeline. Instead, they focused on the compute-heavy training and inference parts of a pipeline.
This created a performance problem, since copying data and operations between CPU-based (say, visualization and feature engineering) and GPU-based (say, model training) steps generates significant overhead from serialization and deserialization compute costs.
RAPIDS handles the whole ML pipeline on GPUs, so anyone can benefit from its speed without having to learn the details of CUDA programming – or any other new tools.