Skip to content
LabPod LabPod LabPod Docs

What is LabPod

LabPod turns a single GPU workstation into a shared platform that a small research team can use from a browser. Each team member gets their own isolated workspace - Jupyter, VS Code, or a plain terminal - with CPU, memory, and GPU quotas plus an advisory disk alert. No SSH setup, no cluster, no scheduler.

The one-liner

Section titled “The one-liner”

Install LabPod on the lab GPU box. Users open http://<host>:24680 in a browser, log in, and launch GPU workspaces. The admin watches who’s using what from a single dashboard and can force-stop any workspace.

Who it’s for

Section titled “Who it’s for”

A 5–10 person research group sharing one or two GPU workstations - a grad-school lab, a small R&D team, a teaching assistant with one powerful machine. The person running the box doesn’t need a Kubernetes background. LabPod installs from a single script and runs as a systemd service.

What you get

Section titled “What you get”
  • Per-user workspaces with enforced CPU/memory/GPU quotas and advisory disk alerts.
  • Browser UI - JupyterLab, VS Code (code-server), a built-in terminal, TensorBoard, MLflow.
  • GPU allocation - whole GPUs, software fractional sharing, or MIG slices (A100/H100).
  • Admin dashboard - live per-user resource usage, GPU occupancy, disk pressure warnings, and audit logs.
  • Rootless container isolation - every workspace runs as the owner’s Linux account; files keep natural host ownership.
  • No queue - coordination is quotas + monitoring + admin force-stop, not a job scheduler.

What it is not

Section titled “What it is not”
  • Not a container GUI - LabPod is a workspace UX, not a general-purpose Portainer replacement.
  • Not a cluster tool - no Slurm, no Kubernetes, no multi-node scheduling.
  • Not a notebook-only tool - workspaces are persistent containers you start and stop, not ephemeral notebook sessions.

LabPod does not auto-stop workspaces. A workspace stays running until you stop it.

Where to go next

Section titled “Where to go next”