---
id: nvidia-nvvm
version: "13.3.73"
license: unclear
license_treatment: unclear
maintenance: active
---

# nvidia-nvvm — NVVM Libraries

License: unclear · Maintenance: active · Downloads: 2.3M/mo

## What it is and what it does

nvidia-nvvm is a Python interface to NVIDIA's NVVM compiler IR library, the intermediate representation layer used to compile CUDA code for GPU execution. It sits between high-level CUDA C/C++ code and final GPU machine code, handling optimization and code generation for NVIDIA hardware.

The package is distributed as platform-specific wheels for Linux (x86_64 and aarch64) and Windows, indicating it bundles compiled binaries. It has no runtime dependencies and is in Beta status, supporting Python 3.5 through 3.11. It is intended for developers, educators, and researchers working on GPU-accelerated systems.

Use it for:

- Building custom CUDA kernels or GPU code generation tools needing direct NVVM IR access
- Developing machine learning frameworks that compile to GPU code
- Creating GPU-accelerated libraries requiring low-level CUDA compilation
- Integrating NVIDIA GPU compilation into Python-based build pipelines

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Provides the NVVM compiler IR library for building and optimizing CUDA applications on NVIDIA GPUs.

Yes, if you are developing CUDA applications or GPU-accelerated software and need programmatic access to NVIDIA's compiler IR. The package is actively maintained, has no known security vulnerabilities, and carries no documented license restrictions. Medium install friction is typical for platform-specific compiled bindings. Not relevant for non-GPU workloads.

## Install

pip install nvidia-nvvm
uv add nvidia-nvvm
poetry add nvidia-nvvm

## Installing nvidia-nvvm

Before you install: Medium install friction due to platform-specific wheels for x86_64 Linux, aarch64 Linux, and Windows. Package is actively maintained with a release 46 days ago and supports current Python versions.

Quickstart:

pip install nvidia-nvvm==13.3.73

import nvidia.nvvm
# Access NVVM compiler IR for CUDA code generation

Requires a compatible platform: x86_64 Linux, aarch64 Linux, or Windows. Python >=3 required.

Verify before relying:

- Whether this package requires a local CUDA toolkit installation or includes all necessary binaries
- Whether the package works on macOS or is limited to Linux and Windows only
- Documentation availability and API stability given Beta development status
- Actual runtime dependencies or system-level prerequisites beyond the wheel distributions

## Package facts

- License: not declared (unclear)
- Python support: supports_current
- Install friction: medium
- Maintenance: active
- Downloads: 2.3M/month (top 5,000 on PyPI)
- Known vulnerabilities: none known

## Tags

cuda compiler ir, nvvm library, nvidia gpu compilation, cuda code generation, gpu compiler backend, nvidia nvvm, cuda optimization, cuda, gpu-compilation

[View on SkillFed](https://skillfed.io/packages/nvidia-nvvm) · [View on PyPI](https://pypi.org/project/nvidia-nvvm/)