NVIDIA/nvidia-container-toolkit resource

Health v2 · maintenance only

Adoption not part of the score

4528 stars · 570 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

The NVIDIA Container Toolkit enables building and running GPU-accelerated containers by providing a container runtime library and utilities that automatically configure containers to leverage NVIDIA GPUs. It works with Docker and other container runtimes on Linux hosts with NVIDIA drivers installed.

Use cases

• run docker containers with nvidia gpu access
• gpu-accelerated machine learning containers
• run cuda workloads in kubernetes pods
• expose host gpu to containerized applications
• set up gpu support for docker on linux
• run deep learning training in containers

When to choose

• you need to run GPU-accelerated containers with Docker, containerd, or Kubernetes on Linux
• you want CUDA workloads in containers without installing the CUDA toolkit on the host
• you are deploying ML/AI inference or training in containerized environments

When to avoid

• you are on Windows or macOS without a Linux host or WSL2 setup
• your workloads do not require NVIDIA GPUs
• you need GPU virtualization or MIG management beyond container runtime configuration

Facets

infra-config · maturity stable

container-runtime gpu-computing developer-tools deployment gpu-computing cloud-computing machine-learning go nvidia gpu cuda containers docker container-runtime kubernetes ai-workloads devops linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="NVIDIA/nvidia-container-toolkit")