GA Capital · Technology & Infrastructure
Data Center IntelligenceCampuses, Capex, Power
Not every AI headline is an investment signal. The useful question is which U.S. hyperscaler and AI-lab campuses clear into disclosed dollars, megawatts, and interconnection reality — and how quickly company filings confirm the spend.
Campus Spend Clears Through Power Before Headlines Settle
Most investable U.S. data-center risk sits at the intersection of named campuses, interconnection reality, and filing-backed capex. Ambient AI coverage without those anchors should stay noise.
Underwrite U.S. data-center risk from campus announcements and filing-backed capex signals together. Headline MW without a state, utility, or operator trail should stay provisional.
Interconnection queues, substation proximity, and utility load programs decide which announced campuses can energize on schedule. Land and tax incentives matter only after power clears.
Hyperscaler guidance, chip-scale disclosures, and named campus spend show whether AI infrastructure budgets are clearing into buildable U.S. sites — or remaining corporate-level intent.
Named Campuses and Capex Filings Set the Investable Map
Underwriting starts where operators name sites, disclose dollars or megawatts, and leave a utility trail. From there, the question is which state footprints and company signals confirm that spend is becoming real.
U.S. data-center investment is not every AI headline. It concentrates where operators name campuses, disclose dollars or megawatts, and commit to utility and construction timelines that can be checked against filings.
This report treats that market as an underwriting problem. It asks which site announcements carry investable location and capital detail, which company infrastructure signals confirm multi-year spend, and which state footprints are densifying fastest under hyperscaler and AI-lab demand.
The live exhibits draw on a seeded research pack of U.S. site announcements and filing-backed company signals, plus a curated wire for fresh power, campus, and capex coverage. They are presented here as an Insights narrative for investment professionals, not as a chat product or Mapbox dashboard clone.
State Footprint Shows Where Announced Capex Is Clustering
Map-visible campuses from the coverage package, with exact city pins separated from state-level proxies. Priority sites rank disclosed investment so the densest capital commitments surface first.
Loading footprint…
Site Announcements Name Where Capital Is Trying to Land
Campus-level rows for announced and expanding U.S. facilities. Prefer operator, geography, status, and disclosed investment over ambient AI infrastructure chatter.
Loading sites…
Company Signals Check Whether Site Lists Match Corporate Spend
Filing-backed capex guidance, chip-scale, and infrastructure program disclosures for major AI and cloud operators. Use these to test whether announced campuses clear into multi-year budgets.
Loading signals…
Curated Wire Filters Ambient AI From Buildable Signal
This feed keeps items that map onto data-center vocabulary — campuses, interconnection, hyperscaler spend, and MW — rather than dumping every AI headline.
Loading wire…
Read Sites and Filings First, Then Confirm With the Wire
Close the report with a simple sequence. Read named campus announcements and company infrastructure signals first, then use the curated wire. AI headlines without a site, utility, or filing trail should stay noise.
Named campuses with operator, geography, status, and disclosed investment or MW where available. These are the primary map of where capital is trying to land.
10-K, earnings, and operator disclosures on capex guidance, chip scale, utility load, and multi-state programs. Use these to check whether site lists match corporate spend.
Fresh coverage of power, interconnection, campus announcements, and hyperscaler buildouts. Ambient AI chatter without a site, utility, or filing trail should stay noise.