revenue per GW of AI capacity; getting 200 MW online 6 months early ≈ $1.2–1.3B gross revenue, ~$400–500M NPV (contested — single-source)
~$12–13B/GW/yrestimatecontested
| Value kind | estimate — Interpret this value according to its displayed kind, scope, source, and as-of date. |
|---|---|
| Caveat | This is the rental/IaaS denominator (SemiAnalysis, contested). Distinct and much larger is the lab token-revenue side: SemiAnalysis's Tokenomics model (Aug 2026) puts OpenAI/Anthropic API inference at >$100B/GW/year on a GB300 cluster against ~$12B/GW/year of rental cost — a model-derived figure sensitive to utilization and price mix, not an audited disclosure. Do not conflate lab API revenue with IaaS rental in one number. |
| As of | 2026-06 |
| Source | SemiAnalysis ('To Boldly Go', Jun 2026) |
| Review | checking…review by 2027-03-29 · fast cadence |
| Recorded changes | last 2026-08-13 · 4 revisions tracked |
| Claim id | revenue-per-gw-of-ai-capacity-1-1-2b-from |
Where the guide uses it
- 2.1 Program & Project Management: The Integrated Master Schedule & Critical Path
- 2.2 Delivery Models & the Owner's Organization
- 3.1 Site Selection Strategy & the Reordered Criteria Hierarchy
- 3.2 Grid Access, Large-Load Service & Speed-to-Power
- 3.3 Power Availability & Power-Cost Structure
- 3.4 Energy Supply Strategy: Grid PPA, BYOP & Co-Location
- 3.5 On-Site & Bring-Your-Own-Power Generation (Energy-Supply Strategy)
- 3.13 Market Clusters & the Site-Scoring Playbook
- 6.6 Construction Execution, Sequencing & Phased Turnover
- 10.5 Provisioning, Bring-Up & Infrastructure as Code
- 13.10 Staged Power/Load Ramp, Go-Live & Handover to Operations
- 14.8 Firmware & Software Lifecycle Management at Fleet Scale
- 15.8 Grid Impact, Energy-Systems Integration & Grid Services
- 16.1 The Power-Bound Era: Why the Bottleneck Moved to the Substation
← Full numbers register — every date-stamped figure in the guide, with revision history.