Power Availability Becomes the Master Constraint on Computing
Grid interconnection queues, not chip supply, now set the timeline for new capacity, and the industry is reorganizing around electrons.
Wikimedia Commons · CC BY-SA 3.0Ask data center developers what gates their next project and the answer is no longer silicon. It is the interconnection queue, the multi-year wait for utility approval to draw large amounts of power from grids that were not planned for computing's growth curve.
The constraint is reorganizing the industry's geography and structure. Development migrates to regions with spare generation and faster approvals; operators contract directly for new generation, sometimes co-developing it; and a market has emerged in sites whose chief asset is a signed interconnection agreement.
On-site and adjacent generation has moved from contingency to strategy. Gas turbines as bridge power, long-term contracts anchoring new renewable buildouts, and serious capital behind small modular nuclear all reflect the same conclusion: waiting for the grid is the expensive option.
The policy dimension is unavoidable. Computing's power demand lands on ratepayers, regulators, and climate commitments simultaneously, and jurisdictions are diverging on who pays for the upgrades and who gets priority in the queue. Where those questions get answered favorably is, increasingly, where the next generation of infrastructure gets built.