Log in to BadgerCompute

At a glance

1. Go to the BadgerCompute Launch URL → 2) Login with your NetID → 3) Approve sign-in using Duo 2FA → 4) Pick a server option → 5) Start your server.

Prerequisites

• You have an active NetID
• You have completed the required BadgerCompute Certification Course
  • Annual renewal may be required.

Step-by-step login

1. Open the BadgerCompute Launch URL
   Visit the BadgerCompute launch page:
   Launch Notebook
2. Click “Sign in with your UW-Madison NetID”
   You’ll be redirected to the UW-Madison Single sign-on login page. Login with your NetID and password.
   
3. Return to BadgerCompute
   After successful SSO, you’ll land back on the JupyterHub page for BadgerCompute.
   
   Two Factor Authentication (2FA)

   You will likely be prompted for 2FA to approve the sign-in via Duo or your chosen method.

   
4. (First time only) Review/accept any terms (if applicable)
   If presented with usage terms and conditions, complete it once. You won’t be asked again unless policies change.
5. Choose your server options
   You may see:
   • Basic Data Science
   • GPU-Accelerated PyTorch
   • GPU-Accelerated TensorFlow
   • Other Docker image (advanced)
   • Build your own image (advanced)
   
   Select the option that fits your workload and click Start.
6. Wait for the server to start
   You’ll see “Your server is starting up…” for ~30–90 seconds (longer during peak times). When ready, you’ll land in JupyterLab.
   
   If it takes too long

   If it takes >3 minutes, refresh the page once. If it still doesn’t start, try a different image. Please note, servers may take longer to start during peak usage times. BadgerCompute is a shared resource and availability is not guaranteed.
7. You’re in! Start working in JupyterLab.
   You can create notebooks, open terminals, and manage files.
   

After you log in

• You’re in JupyterLab. Use the Launcher to open notebooks (.ipynb), terminals, or text editors. See our Welcome to BadgerCompute page for an example data analysis!
• Your files live in your home directory. Anything under your home persists across restarts unless otherwise noted by your site’s storage policy.
• To end your session: File → Log Out (JupyterLab) and optionally Stop My Server on the Hub page to free resources.

Troubleshooting

Sign-out, timeouts, and idle culling

• Sign-out: From JupyterLab, use File → Log Out. On the Hub page, you can also Stop My Server.
• Timeouts: To keep capacity available, servers will be shut down automatically after a period of 4hrs since the start of the session. Files in your /work directory persist for 30 days. All other data may be deleted upon server shutdown.
• Session expiration: Your SSO session will periodically require re-authentication (especially after closing the browser or long idle periods).

Privacy & data notes

• Authentication: You will authenticate via UW-Madison SSO (NetID). Refer to UW-Madison's Privacy Notice for details on how your data is handled.
• Files: Stored in your BadgerCompute home or attached volumes according to the BadgerCompute policy.
• Compute usage tracking: We track aggregate usage metrics to improve the service and plan capacity.

Support

• 💬 Support Forum: community.badgerhub.wisc.edu
• 📚 Docs: docs.badgercompute.wisc.edu