Samsung is investing $1 billion in KKR and Nvidia-backed AI infrastructure firm Helix
Samsung Electronics will contribute $500 million, with five affiliates covering the rest, as Helix targets data centers and power infrastructure
QuartzData center projects, grid demand, power, policy, and capital flows across the U.S. AI infrastructure boom — aggregated from available public sources and proprietary data feeds, with every figure traceable to its source.
Recent data-center headlines · curated, as of Sep 29, 2026.
Samsung Electronics will contribute $500 million, with five affiliates covering the rest, as Helix targets data centers and power infrastructure
QuartzCloud company buys leased facility from landlord
Data Center DynamicsCould cut the amount of cabling in data halls, and speed up deployment times
Data Center DynamicsSamsung Electronics and its five affiliates are investing $1 billion in KKR-launched Helix as the group expands its push into global AI infrastructure.
CNBCWhat's wrong with being a great wholesaler?
The RegisterThe public has soured on AI data centers and states are slowing development, but for HVAC, plumbing, welding, and electrical workers, it's been a jobs boom.
CNBCFederal regulatory actions are allowing on-site generators at data centers and other power plants to release dangerous emissions with fewer checks, according to
Utility Dive“Microsoft claims to want to be a good neighbor, but the jury is still out."
Ars Technicadatacenters.builtfor.ai is a source-aware intelligence platform tracking the buildout of artificial-intelligence data-center infrastructure across the United States. We currently map 1,547 data-center facilities across 46 states and follow the power, policy, and capital shaping where and how AI infrastructure gets built — aggregated from available public sources and proprietary data feeds, and refreshed on a schedule, not hardcoded.
Electricity is the binding constraint on AI data centers. As of Sep 29, 2026, U.S. grid demand was about 436.5 GW, which we track across 7 ISO/RTO markets covering roughly 69% of national load, using U.S. Energy Information Administration (EIA) and CAISO data. The leading generation sources at that time were Natural Gas (44%), Nuclear (18.4%), Coal (16.6%). For the live figure, see the power & grid dashboards.
The geographic footprint is concentrated: the leading states by mapped facilities are Virginia (342), Texas (140), Oregon (133), California (111), Ohio (90). See the full interactive U.S. data-center map and project list.
Data-center siting, power access, water use, and tax incentives are increasingly contested. We track 589 relevant state and federal bills via LegiScan in our policy tracker, refreshed hourly.
The buildout is capital-intensive. We track 93 frontier data centers representing about 13.5 GW and $511.4B in capital expenditure (Epoch AI), and follow 6 public AI-infrastructure equities plus deals, financings, and SEC filings in markets — recently, Nvidia / KKR: Samsung to inject $1 billion into Nvidia- and KKR-backed AI infrastructure firm.
datacenters.builtfor.ai maps 1,547 U.S. data-center facilities across 46 states, sourced from OpenStreetMap. Virginia currently has the most mapped facilities.
As of Sep 29, 2026, U.S. electricity demand was about 436.5 GW, tracked across 7 ISO/RTO markets that cover roughly 69% of national load, using U.S. EIA and CAISO data; the live figure is on the power dashboard. The leading generation sources are Natural Gas (44%), Nuclear (18.4%), Coal (16.6%).
By mapped facilities, the leading states are Virginia (342), Texas (140), Oregon (133), California (111), Ohio (90).
We track 589 state and federal bills relevant to data-center siting, power, water, taxation, and transparency via LegiScan, refreshed hourly.
We follow 6 public AI-infrastructure equities alongside the latest deals and SEC filings — for example, Nvidia / KKR: Samsung to inject $1 billion into Nvidia- and KKR-backed AI infrastructure firm.
Learn how we source and score every figure on our methodology page. This overview is generated from current data and updates as the underlying figures refresh.