How to Connect a Notebook to a Kernel

To connect or switch kernels, open the kernel configuration menu in the top-right of the notebook IDE.

Option	Description	When to use
Running Kernels	Connect to an existing active kernel (e.g. Python 3).	Resume work quickly, save resources.
Start Other Kernels	Launch a new kernel with chosen GPU resources (1×, 2×, 4×, 8× H100).	New experiments or heavy workloads.
No Kernel	Open the notebook without attaching any compute.	View or edit code without running.

Start a new kernel

1. Under Start Other Kernels, select the GPU type you want.
2. Click Confirm and approve the confirmation dialog.
3. Wait 1–2 minutes for the kernel to be ready.
4. Once connected, the new kernel appears under Running Kernels.

Switch to an existing kernel

1. Under Running Kernels, select the kernel you want to connect to.
2. Wait 1–2 minutes for the notebook to be ready.
3. Once connected, the selected kernel is shown under Running Kernels.