Dashboard

Cluster capacity, your quota, and what is running right now.

The landing screen. What it shows depends on your role: an administrator sees the fleet, a user sees their own share of it.

Cluster GPUs by series, fleet GPU-hours, and the submission heatmap.

What you see

Region	What it means
GPUs available	Free cards per series. in use is what running jobs hold; unschedulable is cards the platform has excluded
GPU series	Which series are online. These are the names --profile accepts
My concurrent GPUs	What you are holding against your quota right now
Active jobs	Everything of yours in RUNNING or PENDING
Recent submissions	Your last few, newest first
Submission heatmap	Fleet activity over time, for administrators

Unschedulable is not the same as busy. A card lands there when health checks found pending ECC retirements or uncorrected errors, or when a process outside the platform is holding it — somebody who logged into the node and started a notebook. The capacity view says which.

What you can do

• Click any active job to open its detail view.
• Click a series name to see the profiles built on it.
• Follow Quota and usage for the longer view: usage.

When the numbers look wrong

Symptom	Usually
Free GPUs shown but submissions are refused	A scheduling window is closed, or your quota is exhausted rather than the cluster’s
Cards permanently unschedulable	Failing health checks. The capacity view names the reason; the node needs attention
No series at all	The hardware registry is empty — an administrator configures it in settings