Skip to main content

Google's $17 Billion Wyoming Bet: First Data Center in the State

Big Tech's Nuclear Race: Google, Amazon and Microsoft Bet Atomic Power to Fuel AI Data Centers

Stylized illustration of a nuclear power plant with cooling towers connected by glowing energy lines to AI data center server racks

Illustration: a stylized nuclear power plant feeding glowing energy into rows of AI data center servers.

What Happened

Google and Constellation Energy announced on October 6, 2026 a 20-year agreement supporting 890 megawatts of additional nuclear capacity at 11 reactors in Illinois, Pennsylvania and New Jersey, according to reporting from TokenPost and Android Headlines. Constellation said the arrangement represents more than $4.3 billion in new investment, with the first reactor uprate expected in 2028. In return, Constellation signed a five-year contract to use Google Cloud and Gemini Enterprise tools to manage outage workflows and optimize generator dispatches.

The deal follows a strikingly similar move by Amazon. On September 30, 2026, Amazon and Constellation announced a 20-year power purchase agreement involving 690 megawatts of nuclear power and more than $3 billion in investment at Maryland's Calvert Cliffs nuclear plant, according to Cybernews, with potential for the capacity to expand by another 190 megawatts by the early 2030s.

Microsoft, meanwhile, had partnered earlier with Constellation to restart the historic Three Mile Island nuclear facility in Pennsylvania, and Meta committed in September 2026 to as much as 6.6 gigawatts by 2035 — including a 1.2-gigawatt campus with Oklo in Ohio for its Prometheus AI supercluster, with pre-construction work beginning this year and first power possibly arriving as early as 2030, according to reporting in WebProNews and Utility Dive.

The pattern is unmistakable: the world's largest technology companies are bypassing traditional, slow-moving energy procurement to form direct, long-term partnerships with nuclear operators, guaranteeing the energy security their AI ambitions demand.

Stylized illustration of tech companies racing toward a nuclear power plant finish line as a robot arm builds a data center

Illustration: a stylized race among tech giants toward atomic power for their data centers.

Background

The nuclear push is driven by a simple physics problem. Unlike wind or solar, which are intermittent and depend on weather, nuclear power provides constant "baseload" electricity. AI data centers run around the clock and consume power at rates that dwarf traditional computing, making a steady, weather-independent source essential.

The scale of the demand surge is enormous. According to Gartner data cited by InfotechLead, global data-center electricity consumption is forecast to rise 26.4 percent from 447 terawatt-hours in 2025 to 565 TWh in 2026, while AI-optimized servers alone will consume 175 TWh — an 84.2 percent jump from 95 TWh in 2025 — accounting for roughly 31 percent of total data-center electricity use. Overall consumption could reach 702 TWh in 2027 and exceed 1,200 TWh by 2030, with power demand climbing from 104 gigawatts in 2025 to 132 gigawatts in 2026 and potentially 290 gigawatts by 2030.

Google's own 2026 environmental report said its electricity demand increased 37 percent year over year in 2025. The Constellation deal is only one strand of a broader strategy: in October 2025 Google signed a 25-year agreement with NextEra Energy to restart Iowa's 615-MW Duane Arnold Energy Center, the same year it agreed with Elementl Power to develop three potential US advanced-nuclear sites each targeting at least 600 MW, and in 2024 it backed Kairos Power's plan to deploy an advanced reactor fleet capable of producing up to 500 MW by 2035, according to Cybernews. Google has also pursued nuclear power in Finland through a deal with Fortum.

But the race is creating friction with the grid. On October 7, Duke Energy announced a proposed settlement in North Carolina that would require new customers using at least 50 megawatts, with an 80 percent load factor, to pay upfront for dedicated facilities and provide deposits or guarantees for shared upgrades, for electric-service agreements signed after June 1, 2026 — a decision expected by mid-November, TokenPost reported. Grid operator PJM has responded by supporting voluntary "Bring Your Own New Generation" arrangements that let large loads bring new power to offset their demand. Amazon, Google and other operators support paying for dedicated facilities but resist automatically footing the bill for shared grid upgrades tied to their new demand.

Why It Matters

The nuclear deals reframe the AI boom as an energy story. Hyperscalers can secure GPUs, networking gear and land for new data centers, but those investments have limited value without sufficient electricity — making reliable power a binding constraint on how fast AI can grow. By locking in 20-year supplies, tech giants are effectively reserving a share of the US power system years before the reactors are uprated.

There are climate implications too. Nuclear generation produces no direct carbon emissions, which helps companies like Google, whose AI expansion drove carbon emissions up 18 percent last year alongside the 37 percent leap in power demand, meet climate commitments. Amanda Peterson Corio, Google's Global Head of Data Center Energy, has noted that squeezing extra megawatts from existing nuclear plants offers the fastest, most cost-effective path toward cleaner grid capacity — a significant admission that building brand-new reactors takes a decade or longer to permit and construct.

For consumers and local communities, the open question is who pays for grid upgrades. Utilities argue that the massive new loads should cover the cost of reinforcing the network; tech companies argue that blanket cost-sharing would penalize efficient operators. The North Carolina settlement and PJM's voluntary framework could become templates for resolving the fight nationwide — with household electricity bills potentially in the balance.

Stylized illustration of giant server racks shaped like brains straining a power grid while a nuclear cooling tower glows steadily behind

Illustration: a stylized depiction of AI data centers straining the power grid while nuclear energy stands by.

What Happens Next

  • Constellation's first reactor uprate under the Google deal is expected in 2028, with further uprates across the 11 reactors to follow.
  • The North Carolina Utilities Commission is expected to rule by mid-November on Duke Energy's proposed upfront-payment terms for large new electricity customers.
  • PJM's voluntary "Bring Your Own New Generation" framework will be tested as more data centers connect in its territory.
  • Meta's Oklo partnership and Blue Energy's Texas small-reactor project will show whether advanced fission can move from announcements to actual construction.
  • Investors will track how the nuclear spending spree reprices utility assets — and whether the $4.3 billion-plus Google commitment sets the going rate for baseload AI power.
  • Regulators and consumer advocates will keep pressing on the question of who funds shared grid upgrades as AI loads keep growing.

Sponsored: explore today's featured sponsored offers

FAQ

Q: What is the Google–Constellation nuclear deal?

A: Announced October 6, 2026, it is a 20-year agreement for 890 megawatts of additional nuclear capacity at 11 reactors in Illinois, Pennsylvania and New Jersey, representing more than $4.3 billion in investment, with the first reactor uprate expected in 2028.

Q: What did Amazon and Microsoft do?

A: Amazon announced on September 30, 2026 a 20-year agreement with Constellation for 690 MW at Maryland's Calvert Cliffs plant plus over $3 billion in investment; Microsoft partnered with Constellation to restart the Three Mile Island facility.

Q: Why are tech companies choosing nuclear power?

A: AI data centers run around the clock and need constant "baseload" electricity. Nuclear provides steady, emissions-free power that, unlike wind or solar, does not depend on weather.

Q: How fast is AI power demand growing?

A: Gartner data puts global data-center electricity at 565 TWh in 2026, up 26.4 percent from 2025, with AI-optimized servers consuming 175 TWh — 84.2 percent more than in 2025 — and total demand potentially exceeding 1,200 TWh by 2030.

Q: Who pays for the grid upgrades these data centers need?

A: That is contested. Utilities want large users to pay upfront for dedicated facilities and guarantee shared upgrades; tech companies support dedicated facilities but resist automatic cost-sharing. A North Carolina proposal expected to be decided by mid-November may set a template.

Sources

  • Cybernews (Oct 6, 2026) — Google's nuclear deal with Constellation; Amazon's 690 MW Calvert Cliffs agreement; Google's broader nuclear strategy.
  • Android Headlines (Oct 7, 2026) — Google's billion-dollar-scale Constellation deal for 890 MW of new nuclear capacity.
  • TokenPost (Oct 13, 2026) — Google–Constellation 20-year, $4.3B agreement details; Duke Energy settlement; PJM grid arrangements.
  • InfotechLead (Sep 2026) — Gartner data-center electricity forecast: 565 TWh in 2026, AI servers at 175 TWh, up to 1,200 TWh by 2030.
  • WebProNews / Utility Dive (Sep 2026) — Meta's up-to-6.6 GW nuclear commitments including the 1.2 GW Oklo Ohio campus.

Comments

Popular posts from this blog

CERN's LHCb Finds Evidence of One-in-Ten-Billion Particle Decay After 42-Year Search

GENEVA — Physicists working at CERN's Large Hadron Collider have found the first evidence from a single experiment of one of nature's rarest particle decays, a process so unlikely that the Standard Model of particle physics predicts it should happen only once in every ten billion beauty particles. The result, announced at a seminar at CERN on Tuesday, November 3, 2026, ends a search that began in the early 1980s with the CLEO experiment at Cornell University in the United States. It comes from the Large Hadron Collider beauty experiment, known as LHCb, which studies particles containing beauty quarks — heavier relatives of the quarks that make up much of the visible matter in the universe. The decay in question involves a neutral B0 meson, a short-lived particle made of a beauty quark and a down antiquark, transforming into a pair of muons, heavier cousins of the electron. According to the University of Birmingham, which plays a leading role in the LHCb collaboration, the mea...

Oil Plunges Below $100 as Markets Rally After US Midterm Sweep

LONDON — Global markets delivered a dramatic split verdict on Thursday, 5 November 2026, as oil plunged below $100 a barrel while shares, gold and most major currencies rallied after Democrats swept the US midterm elections. Brent crude fell 97 cents, or 1.03 percent, to $93.65 a barrel as of 0211 GMT, while US West Texas Intermediate dropped $1.14, or 1.29 percent, to $86.98, according to Reuters. Both benchmarks had tumbled more than 8 percent in the previous session to their lowest since 31 August, and Brent is about 20 percent lower this week, heading for its biggest weekly drop since the pandemic shock of April 2020. Yet the S&P 500 closed 1.6 percent higher on Wednesday, European shares neared a two-month high, and gold hit its highest in more than two weeks. Investors are pricing in a lower chance of renewed US-Israeli strikes on Iran and a greater chance that the war which began at the end of February 2026 ends through negotiation. Illustration: An oil pumpjack beside ...

Wall Street's Fear Gauge Soars to Eight-Month High as Nasdaq Slides Into Correction on AI Selloff

NEW YORK — Wall Street's most closely watched gauge of market anxiety logged its biggest jump in more than 18 months on Thursday, 5 November 2026, as a brutal selloff in chip and data-center stocks dragged the Nasdaq Composite more than 10% below its record closing high and into correction territory. The Cboe Volatility Index, known as the VIX, surged 11.5 points to an eight-month high of 28.09, its largest one-day rise since the days immediately after US President Donald Trump unveiled sweeping import tariffs on "Liberation Day" in April 2025 — a violent reversal of the post-election rally that had lifted markets just a day earlier. Illustration: The New York Stock Exchange on Wall Street, where the Nasdaq slid into correction territory as the VIX spiked to an eight-month high. What Happened The selloff struck fast and hit the market's most crowded trade hardest. The Nasdaq Composite fell 4% on Thursday, according to Reuters, after falling nearly 4% to 24,712....