How much compute is your
cluster wasting?

One command. 30 seconds. See exactly where resources go to waste.

$ curl -s https://wastage.expanse.sh/scan -o scan.sh && bash scan.sh

Works on SLURM and Kubernetes. All processing happens locally on your machine.

538.8k

jobs analysed

95.9M

core-hours wasted

$9.6M

estimated waste

2

clusters scanned

How it works

1. Run the command

The script auto-detects your scheduler (SLURM or Kubernetes), collects resource usage data, and computes waste metrics entirely on your machine. No data leaves your cluster.

2. See your waste report

CPU, memory, and GPU waste breakdown with estimated dollar cost. A utilisation score from 0-100 tells you how efficiently your cluster runs.

3. Share with your team

Get a shareable URL with a visual report you can send to your manager, post in Slack, or share on Twitter.

Leaderboard

Ranked by utilisation score. Opt-in only.

#ClusterScoreSchedulerCountry
1Cirrus59/100SLURMScotland
2ARCHER241/100SLURMScotland

Want more than a one-time scan?

Expanse installs on your cluster in one command and tracks CPU, memory, and GPU waste continuously, with per-job breakdowns and historical trends. Free.

Install Expanse free →