GPUs and notebooks for your classroom
Give students a hosted JupyterHub, a browser IDE, modern GPUs, and frontier open-weight models, without asking them to learn Kubernetes or asking your department to buy hardware. Free to U.S. nonprofit institutions, community colleges included.
What you get
Everything below is already running. You do not have to stand any of it up yourself.
Point students at a URL and they get a notebook server. Nothing to install, no per-student setup, no cluster knowledge required.
Need a specific image, package set, or resource limit for your course? We will help you run your own hub with the environment your syllabus assumes.
A hosted VS Code in the browser, so a lab assignment is a link rather than an afternoon of environment troubleshooting.
Real GPUs for machine learning, data science, and HPC coursework: enough for a full class to train at once, not one shared card.
A catalog of frontier open-weight models behind an OpenAI-compatible endpoint. Students can build against an API without a credit card or a rate limit surprise.
Bi-weekly office hours with the NRP team and recorded trainings on Docker, Kubernetes, and JupyterHub, for you and for your students.
Bringing a course to the NRP?
Tell us the course, roughly how many students, and when it starts. We will get the namespace and resources in place before your first lecture.