Inside the U.S. Hyperscaler Buildout: A Cycle Still Understated

The U.S. is the anchor point of global AI and cloud infrastructure. Yet, as hyperscalers signal plans to double their capacity by 2027, questions remain around what that growth looks like in physical terms.

Aterio’s latest dataset helps answer this: identifying the number of active, under-construction, and announced data centers across the country, along with a substantial set of undisclosed projects, sites not yet publicly linked to a provider but due to scale we can assume that most of that activity is going to be for hyperscalers.

Our data lets us quantify this surge, analyzing where the largest expansions are taking place and which states are absorbing the bulk of hyperscaler-led investment in land, power, and construction.

Current Footprint vs. Future Pipeline

To understand the scale of this expansion, we start by looking at the current footprint of hyperscaler-owned data centers across the U.S. and how that compares to the pipeline of new developments.

Aterio’s dataset shows that the “Big Four” providers: Amazon AWS, Microsoft, Google, and Meta, operate nearly 500 self-built facilities nationwide, with more than 530 additional buildings announced or under construction. Within that total, AWS leads with roughly 197 active data centers, followed by Microsoft (102), Meta (93), and Google (82).

It’s important to note that not all announced projects are guaranteed to materialize, some still depend on local approvals or power availability. Even so, the scale of the pipeline indicates a near-doubling of existing capacity, underscoring the momentum behind the current buildout cycle and its potential to reshape U.S. digital infrastructure.

The Undisclosed Pipeline

Even as hyperscalers confirm record expansion plans, Aterio’s data shows an even larger layer of announced but undisclosed projects moving through early development stages. These total more than 1,000 planned data center buildings across the U.S., many exhibiting the same physical and power characteristics as hyperscaler sites.

While ownership has yet to be confirmed, the scale and geographic distribution of these projects suggest that hyperscalers are driving much of this unreported activity. However, the scale of this cycle remains understated by most market estimates, indicating that the real extent of U.S. data center expansion may far exceed what is currently visible in public disclosures.

Geographic Concentration of Planned Projects

The next layer of the analysis looks at where this expansion is happening. Aterio’s latest data shows that planned capacity is highly concentrated across a few key states: Virginia, Texas, and Georgia, which together account for nearly half of all new U.S. data center developments.

Virginia remains the dominant hub, driven by sustained investment in Loudoun County’s power and fiber corridors. Texas follows with large-scale campuses planned around Fort Stockton, Abilene, and Amarillo, supported by favorable land availability and new gas and transmission infrastructure. Georgia, Pennsylvania, and Arizona round out the next tier, each benefiting from expanding energy grids and pro-development policy frameworks.

This geographic clustering underscores how infrastructure, regulation, and power access are shaping where the next phase of hyperscaler growth occurs and which local markets are positioned to absorb the majority of upcoming investment.

Takeaway for the Current Cycle

The U.S. hyperscaler buildout continues to accelerate, with confirmed projects already signaling a near-doubling of capacity. The undisclosed pipeline adds another layer of potential growth, underscoring how the true scale of this cycle remains understated by most market estimates.

As approvals advance and power infrastructure expands, much of today’s unconfirmed capacity will likely move into active construction, and we will be monitoring closely.

Access Aterio’s latest U.S. Data Center Report to explore detailed breakdowns by state, provider, and development stage, or book a call to see how our data can support your investment analysis.