NVIDIA/libnvidia-container

NVIDIA container runtime library

What it solves

It simplifies the process of configuring GNU/Linux containers to use NVIDIA GPUs. It removes the need for manual configuration of hardware access and driver libraries within a containerized environment.

How it works

The project provides a library and a command-line interface (CLI) utility that leverages kernel primitives to automatically configure the container's root filesystem. It is designed to be agnostic of the specific container runtime being used.

Who it’s for

Developers and system administrators who need to expose NVIDIA hardware to Linux containers regardless of the container runtime.

Highlights

  • Automatically configures NVIDIA hardware for containers.
  • Runtime agnostic implementation.
  • Includes a CLI utility for direct configuration.
  • Supports multiple Linux distributions including Ubuntu, Debian, CentOS, and Amazon Linux.

Related

  • Project
    NVIDIA/gpu-operator

    An automation tool for Kubernetes that manages the installation and lifecycle of NVIDIA GPU drivers, runtimes, and monitoring tools as containers.

  • Project
    NVIDIA/open-gpu-kernel-modules

    The source release of NVIDIA's open GPU kernel modules for Linux, enabling the build and installation of kernel-level driver components for Turing and newer GPUs.

  • Project
    NVIDIA/k8s-device-plugin

    The NVIDIA k8s‑device‑plugin is an official DaemonSet that implements Kubernetes’ device‑plugin API for NVIDIA GPUs. It discovers GPUs, reports health, and makes them schedulable via the `nvidia.com/gpu` resource (with optional MIG and shared‑GPU modes). Install the NVIDIA driver and container toolkit on each node, deploy the plugin (static manifest or Helm), then request GPUs in pod specs. Configuration is possible via flags, env‑vars, or a YAML file, covering MIG strategy, device‑list delivery, shared‑GPU (time‑slicing or MPS), and more. The plugin is beta, lacks full health‑check/cleanup, and only the official NVIDIA version is supported.

  • Project
    basetenlabs/truss

    A CLI tool for packaging and deploying AI/ML models to production, automating containerization, GPU configuration, and dependency management.

  • Project
    mmartial/ComfyUI-Nvidia-Docker

    Docker images that run the ComfyUI Stable‑Diffusion web UI with NVIDIA GPU acceleration, offering GPU‑specific tags, UID/GID mapping, and built‑in ComfyUI‑Manager for easy updates.