What is this?
nvidia-docker was a tool that let Docker containers access NVIDIA GPUs. It has been deprecated and the repository archived, replaced by the NVIDIA Container Toolkit.
Is this a library or a command-line wrapper?
It was a command-line wrapper around Docker that configured GPU access, not a library you import into code.
What should I use instead of this?
Use the NVIDIA Container Toolkit, which extends Docker to use the NVIDIA Container Runtime; follow its documentation and install guide instead.
Should I still install nvidia-docker today?
No. The repository is archived and the wrapper is no longer supported or maintained.
What is the NVIDIA Container Toolkit?
It is the successor project to nvidia-docker; it configures Docker to use the NVIDIA Container Runtime so containers can access GPUs.
Can I still use an existing nvidia-docker install?
It will likely still function but receives no updates or support, so migrating to the NVIDIA Container Toolkit is recommended.
What are similar projects?
Soup, olmocr, InstinctFlash, GLM-5.3-Flash-EXL3-2x-DGX-Sparks-TensorFold, smolvm, TornadoVM, LoongForge, nvitop
Which topics is it in?
containers, gpu-acceleration, nvidia, deprecated, cuda, devops, go, command-line-tool, deep-learning, gpu, machine-learning, apache-2.0