Layer 1 · How the heat leaves
Air cooling
General compute and any AI deployment under about 30 kW a rack.
The shape of it
- Capital cost
- Lowest
- Maturity
- proven
- Density supported
- Up to ~30 kW per rack
- PUE
- 1.4-1.6 typical
Verified 2026-09-09.
The catch
It runs out. The TCO crossover where liquid wins sits near 30 kW a rack at $0.12/kWh, and the thermal inflection near 45-50 kW. Above that, air is not a cheaper choice — it is not a choice.
What it actually decides
Still correct for most racks in most buildings. Simply not for the ones running frontier accelerators.
What this is powering
A GB200 NVL72 rack draws 120–132 kW and a GB300 pushes 135–200 kW, against a 2026 average rack of about 27 kW. The silicon decision and the power decision are the same decision, made eighteen months apart.
Layer 2 — Silicon · GB200 NVL72 · The 150 W inference option
Weighed against
- Air cooling vs Direct-to-chip liquid cooling — Where the crossover actually sits: about 30 kW a rack on cost, 45-50 kW on thermals.
Others in how the heat leaves
- Rear-door heat exchangers — ~30-50 kW per rack
- Direct-to-chip liquid cooling — 50-200+ kW per rack
- Single-phase immersion — 100+ kW per rack equivalent
- Two-phase immersion — Highest available
- Free and evaporative cooling — Depends on the paired system
Sources
Want this whole layer as one table?
The power and cooling options on these pages as a single CSV — lead time, capex, carbon, maturity, density supported, PUE, the catch and what each one decides — ready to sort, filter or load into a notebook.
Get the energy table — free for research and evaluation
Four short lines and the download starts at once (CC-BY-NC-4.0, built from these pages the moment you ask). What you type is used only to answer the request and is never sold. Prefer Hugging Face? Request it there.
Using it inside a product, a client report or anything you sell? That needs a commercial licence.