Megawatts to Megacampuses
Site Selection, Power Strategy, and the $3 Trillion AI Infrastructure Build-Out
Description
The demand for AI compute has triggered the most capital-intensive real estate build-out in modern history — and the gap between capital committed and capacity actually delivered has never been wider. Power access has become the primary constraint on development decisions, reshaping where projects get built, how sites get selected, and how hyperscalers structure their relationships with developers. Primary markets are running out of room. Secondary markets are absorbing unprecedented capital without established playbooks. And the decisions being made right now — on site selection, power procurement, build vs. lease, and market expansion — will define the geography of AI infrastructure for the next decade. This roundtable convenes senior real estate developers and hyperscaler infrastructure leaders for an honest peer exchange on what it actually takes to turn capital conviction into delivered capacity at scale.
Background
The five largest tech companies will spend up to $700 billion on U.S. data centers in 2026 alone. Yet data center vacancy has fallen below 2%, grid interconnection timelines stretch 4–10 years in primary markets, and construction costs are rising 7% annually. Demand has never been stronger — and the gap between capital committed and capacity delivered has never been wider. Developers are racing in…
Date: 2026-07-16
Time (ET): 2:00 PM EDT, Jul 16, 2026
Time (Local): 6:00 PM UTC, Jul 16, 2026
Location: online