What is the difference between CUDA-X AI and RAPIDS?

Hi all, what is the difference between CUDA-X AI and RAPIDS?, I see both are oriented to accelerate the complete end-to-end data science pipeline throw a collection of software libraries (most of them are common to both SDK such as cuDF, cuML, cuGRAPH, cuDNN).

RAPIDS suite is a data science ecosystem that can be used as a single unified tool via its containerized version https://ngc.nvidia.com/catalog/containers/nvidia:rapidsai:rapidsai. What about CUDA-X AI?

I read that “CUDA-X AI is a collection of software-acceleration libraries built on top of CUDA. Its software-acceleration libraries are integrated into all deep learning frameworks, including TensorFlow, PyTorch, and MXNet, and popular data science software such as RAPIDS”. Does it mean that RAPIDS is built on top of CUDA-X instead of the traditional version on CUDA?