ai-opsLive
GPU fleet overview
GPU fleet
Utilization across the inference cluster.
GPU utilization
78%+4pt
Nodes online
24/240
Queue depth
12-3
Avg. queue time
1.2s-0.3s
Node pools
inference-pool-a (A100)
8 nodes · 82% util
inference-pool-b (H100)
12 nodes · 74% util
training-pool (H100)
4 nodes · 91% util