NVIDIA provides documentation on how to program their GPUs in Python for their CUDA architecture.

Python plays a key role within the science, engineering, data analytics, and deep learning application ecosystem. NVIDIA has long been committed to helping the Python ecosystem leverage the accelerated massively parallel performance of GPUs to deliver standardized libraries, tools, and applications. Today, we’re introducing another step towards simplification of the developer experience with improved Python code portability and compatibility.

Our goal is to help unify the Python CUDA ecosystem with a single standard set of low-level interfaces, providing full coverage of and access to the CUDA host APIs from Python. We want to provide an ecosystem foundation to allow interoperability among different accelerated libraries. Most importantly, it should be easy for Python developers to use NVIDIA GPUs.

The documentation goes over the code and steps needed to use Python with CUDA. The performance is especially impressive for Python vs. C++:

See the article here.