bwJupyter: Your Cloud-Based Laboratory for Data Science & AI
bwJupyter is a powerful, web-based development environment that was specially developed for students and teachers in Baden-Württemberg. It allows you to create interactive documents (notebooks) that combine live code, equations, and rich media – without having to install software on your own computer. By offloading the computing load to the country's high-performance computing clusters, bwJupyter ensures that you can run complex simulations and data visualizations directly in the browser. This allows for a seamless and fair learning experience for everyone, regardless of the hardware they have.
Beyond classic programming, the platform is a gateway to state-of-the-art technologies. It offers free access to NVIDIA professional GPUs, making it an indispensable tool for training AI models and machine learning. To further support the learning process, the environment includes AI-powered coding wizards that help debug, explain, and optimize scripts in real time. Whether in a seminar or workshop: Simply log in with your university entrance data (Shibboleth) and start your next digital project immediately.
Why use it in teaching?
- No installation: No "it doesn't work on my machine" issues – everyone uses the same pre-configured environment.
- AI-supported: Native AI tools help students learn Python, R, or Julia more intuitively.
- High performance: Free access to professional GPU power for AI and big data projects.
- Easy access: One-click sign-up via the standard university account.
bwJupyter: Latest Updates and New Features
In the winter semester 2026, bwJupyter continues to support universities and teaching staff in implementing digital teaching. The service is continuously being developed and expanded with new features and resources.
The following updates and enhancements have been introduced since the summer semester 2026:
More Open Educational Resources through CAMMP
CAMMP (Computational and Mathematical Modeling Program) is an inter-university project that develops, among other things, digital workshops on mathematical modelling.
More than 40 of these workshops are now available as Jupyter Notebooks on bwJupyter, where they can be explored and edited.
Copying Profiles
As part of the regular deprovisioning process, profiles and data that are no longer needed are deleted.
To make it easier to recreate profiles, users can now copy existing profiles, including their contents.
Providing Git Repositories
Teaching staff can now make content from existing GitHub repositories directly available on bwJupyter. When creating a profile, a repository can be specified and will be cloned automatically. This eliminates the need to upload the content again to a shared folder.
Local Guest Accounts
Local guest accounts are now available on bwJupyter for events such as workshops and seminars. These accounts function like regular JupyterHub accounts but allow users to log in with a username and password instead of using bwIDM.
After receiving the necessary approval, teaching staff can create multiple guest accounts with a limited validity period. These accounts are automatically deleted once their validity expires.
A complete overview of all updates and enhancements is available in the bwJupyter changelog.
bwJupyter Community Meeting on 12 October
The third bwJupyter Community Meeting will take place on Monday, 12 October 2026, at 3:00 p.m.
During the meeting, selected new features will be presented. Participants will also have the opportunity to ask questions and provide feedback, helping to shape the future development of bwJupyter.
The event will take place online via Zoom.
Date: Monday, 12 October 2026, 3:00 p.m.
Location: Zoom
Access: kit-lecture.zoom-x.de/j/61082602332
