"Python is a “first-class” language at NVIDIA CUDA, and NVIDIA is committed to bringing our technology to Python developers in close alignment with C++ upon new releases of our hardware. We're also happy to announce the general availability of CUDA Python 1.0. NVIDIA's commitment to Python goes well beyond just our own tech stack. NVIDIA's Python engineers contribute across a broad swath of the Python ecosystem, from the core interpreter itself, to packaging and PyPI, to the Python community at large."
"Since PyCon last year, NVIDIA Pythonistas - in collaboration with many others in the Python community - have made great progress on the evolution of various packaging standards, including working with community partners on the implementation of wheel variants and the establishment of a Packaging Council to better govern the evolution of packaging standards and PyPI. NVIDIA Python engineers are also engaged in implementation, testing, and porting work for the free-threaded build of the interpreter."
"NVIDIA Python engineers are driving the early exploratory work for adopting Rust for CPython, work on Python performance benchmarking, and are actively involved in many enhancements for Python 3.14 and 3.15, including providing built-in Zstandard support in Python 3.14. At NVIDIA, we are excited to work with our partners and the open source Python community to help bring the best developer experience for users of high performance computing and AI."
"NVIDIA is excited to once again support PyCon US 2026 as a Visionary Sponsor, and to sponsor the Future of AI with Python Conference Track. Come see NVIDIA at the Anaconda and PyTorch booths, and at the AI Track."
NVIDIA will support PyCon US 2026 as a Visionary Sponsor and sponsor the Future of AI with Python Conference Track. Python is treated as a first-class language at NVIDIA CUDA, with technology aligned to C++ upon new hardware releases. CUDA Python 1.0 is available generally. NVIDIA’s Python engineers contribute across the Python ecosystem, including the core interpreter, packaging, and PyPI, and participate in the broader Python community. Progress includes work on wheel variants, a Packaging Council, free-threaded interpreter build efforts, exploratory adoption of Rust for CPython, performance benchmarking, and enhancements for Python 3.14 and 3.15 such as built-in Zstandard support. NVIDIA plans to collaborate with partners and the open source community to improve developer experience for high performance computing and AI users.
Read at Pycon
Unable to calculate read time
Collection
[
|
...
]