The AI buildout now waits on the grid, not on demand. A site can be permitted and poured long before it can be energized, and much of the world cannot deliver the power these designs assume. Capacity is the constraint that decides where AI gets built next, and it rewards the operators who lined up supply years ago.
- In parts of the US, a project can wait five years just to connect to the grid.
- AI workloads run around the clock and do not shed load at peak.
- That makes it a bad match for a system built for slow, predictable growth.
Full breakdown. AI Data Centers Are Running Out of Power Before They Run Out of Demand
Data centre power is the other half of this story. The 2026 State of Enterprise Infrastructure report covers where the megawatts are going and why grid access gates new capacity.
Get the next one before it is old news
Independent analysis of cloud-native infrastructure, Kubernetes and data centre economics. No vendor spin.