Hyperscalers in 2026: The Sprint Meets the Power Bottleneck
COMPILED NOTES
Power is the binding constraint; 2026 capex >$600B; 770 facility pipeline; 36+ projects ($162B) stalled; Meta >10GW target
Hyperscalers 2026 — The Power Bottleneck
Core Thesis
The 2026 hyperscale sprint is gated by power availability, not capital or chips.
Key Numbers
- Hyperscale capacity expected to double in ~12 quarters.
- 2026 capex projected to exceed $600B.
- 770 future hyperscale facilities in the pipeline; 1,297 operational as of late 2025.
- 36+ projects ($162B) blocked or significantly delayed.
- Hyperscalers need tens of GW of aggregate power over the next 2–3 years.
Sites & Players
- Meta: >10 GW total target by end-2026; Louisiana Hyperion ($27B, 2.5M sq ft, 2→5 GW); Ohio Prometheus 1 GW supercluster (online 2026).
- Oracle/Stargate I (Abilene, TX): 1.2 GW.
- Microsoft: Atlanta Fairwater (operational Oct 2025); Wisconsin (early 2026).
- AWS: Saudi ($5.3B), Chile (>$4B), Germany (€7.8B through 2040).
Binding Constraints
Power grids strained; long-lead equipment (high-voltage transformers, medium-voltage switchgear) sets energization schedules; "outages increasingly common."
Source: Hyperscalers in 2026 — Data Center Knowledge, 2026-01-06. Full-text fetched.
RELATED · IN THE BASE