What fits in my power budget?
Three models on H200 alone: 41 GPUs, 23.6 kW. Add a GB300 NVL72 cluster and the Capacity Planner moves the two big models there: 15 GPUs, 17.8 kW, and a lower bill. A third more work from every megawatt.
Capacity Planner / demo 6
01What you enter, what you get
The three-model portfolio of demo 4 (Llama-3-70B, Qwen3-32B, Qwen3-8B; 130 requests a second), optimized for the lowest energy. Two sites: H200 only, and H200 plus a GB300 NVL72 cluster.
H200 only: 41 GPUs, 23.6 kW, about $73,800 a month. H200 + GB300: Llama-3-70B and Qwen3-32B move to GB300 (8 + 6 GPUs), Qwen3-8B stays on one H200 — 15 GPUs, 17.8 kW (25% less), about $57,240 a month (22% less). Lowest energy and lowest cost pick the same placement here. Every promise met.
02The Capacity Planner's report
Every run produces a report you can share: the workload, the recommended cluster, each traffic pattern against its speed promise, and the economics. This is the summary report from this demo, exactly as the Capacity Planner produced it.
03Watch
The demo in 33 seconds.
The real Capacity Planner screens, sped up.