nvidia-nvfatbin

What it is and what it does

nvidia-nvfatbin is a Python interface to NVIDIA's fatbin compiler library, used for interacting with CUDA binary formats. Fatbin files are containers for GPU-compiled code and metadata; this package allows Python developers to work with these artifacts programmatically, likely for tasks like inspecting, manipulating, or embedding GPU binaries in larger applications.

The package targets developers working with CUDA, machine learning frameworks, and GPU computing. It has no declared runtime dependencies and is distributed as platform-specific wheels for Linux (x86_64 and aarch64) and Windows. The library is actively maintained and has seen recent releases, though its documentation and licensing terms are not clearly published.

Use it for:

• Inspect or extract metadata from CUDA fatbin files in a Python workflow.
• Programmatically manipulate GPU binary artifacts during custom build or deployment pipelines.
• Integrate CUDA binary handling into machine learning or deep learning framework extensions.
• Automate GPU code packaging or versioning tasks that require fatbin introspection.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Provides Python bindings to NVIDIA's fatbin compiler library for working with CUDA binary formats and GPU code compilation artifacts.

Yes, if you need to work with CUDA fatbin files in Python and are on a supported platform (Linux x86_64/aarch64 or Windows x64). The package is actively maintained and has no known vulnerabilities. However, verify the license terms before use in proprietary projects, and confirm that your system meets any undocumented CUDA runtime dependencies. Medium install friction is typical for NVIDIA GPU libraries.

Install

nvidia-nvfatbin on PyPI

pip install nvidia-nvfatbin

uv add nvidia-nvfatbin

poetry add nvidia-nvfatbin

Installing nvidia-nvfatbin

Before you install

Medium install friction due to platform-specific wheels (x86_64 Linux, aarch64 Linux, Windows). Last release 80 days ago with active maintenance status, though repository metadata is not publicly available.

License in practice

License treatment is unclear—no SPDX identifier or raw license text is available in the package metadata. Verify licensing terms before use in proprietary or redistributed projects.

Quickstart

pip install nvidia-nvfatbin==13.3.29
import nvidia.nvfatbin

Requires Python 3 or later; platform-specific wheels available only for x86_64/aarch64 Linux and Windows x64.

Verify before relying

• What specific fatbin operations or APIs does this library expose—is it a thin wrapper or a full compiler interface?
• Are there undocumented system dependencies (e.g., CUDA Toolkit, driver versions) required at runtime?
• What is the actual license under which this package is distributed?

Package facts

Evidence: nvidia_nvfatbin-13.3.29-py3-none-manylinux2010_x86_64.manylinux_2_12_x86_64.whl; nvidia_nvfatbin-13.3.29-py3-none-manylinux2014_aarch64.manylinux_2_17_aarch64.whl; nvidia_nvfatbin-13.3.29-py3-none-win_amd64.whl

Keywords: cuda, nvidia, runtime, machine learning, deep learning

Tags

More Software Development packages