X-Energy’s $1.02B IPO Signals Nuclear Power’s Pivot From Legacy Asset to AI Infrastructure
Amazon-backed reactor developer prices shares above range as tech giants race to secure baseload capacity for hyperscale compute demands.
X-Energy raised $1.02 billion in its April 23 IPO at $23 per share—39% above the marketed range—as Amazon and institutional investors bet that advanced nuclear reactors have become critical infrastructure for AI scaling rather than legacy energy assets.
The oversubscribed offering, which saw double-digit demand from long-only and energy-dedicated funds according to Street Insider, reflects a structural realignment in energy markets. Small modular reactors (SMRs) that once struggled for commercial viability now attract premium valuations as hyperscalers confront a stark reality: grid constraints and intermittent renewable power cannot sustain 24/7 AI compute demands projected to consume 1,000+ terawatt-hours annually by 2030.
$23/share
$1.02B
+39%
Tech Giants Rewire Energy Strategy
Amazon’s backing of X-Energy follows a pattern established across big tech in 2024-2025. Microsoft committed $16 billion to restart Three Mile Island’s 837 MW reactor by 2028. Google signed agreements for 500 MW of Kairos Power SMRs. Amazon itself inked a 1.9 GW power purchase agreement with Talen Energy’s Susquehanna nuclear plant in Pennsylvania, ramping to full volume by 2032, per ESG Today. Collectively, these firms have contracted over 10 GW of new nuclear capacity in twelve months.
The shift reflects recognition that data center electricity demand will surge from 460 TWh in 2024 to potentially 1,300 TWh by 2035, according to projections cited by Introl. Wind and solar cannot meet this load profile—AI training clusters require uninterrupted baseload power. Nuclear’s capacity factor averages 93% versus 25% for solar and 35% for wind.
“Data Centers have grown in size and AI is dramatically changing the future forecast. In the 2030s, the grid will have less coal and there will be some constraints on gas. So Nuclear Energy’s power density and carbon-free high reliability is attractive, and tech companies are starting to take action on new nuclear deployments.”
— Dan Stout, Founder, Advanced Nuclear Advisors
Grid saturation is already measurable. BloombergNEF projects data center power demand in the PJM grid—which serves 65 million people across 13 states—will hit 31 GW by 2030, nearly matching the 28.7 GW of total new generation expected over the same period. Data centers are forecast to consume 6.7-12% of U.S. electricity by 2028, up from 4.4% in 2023. Gartner predicts power shortages will constrain 40% of AI data centers by 2027.
Regulatory Acceleration Enables First-Mover Advantage
X-Energy’s commercial timeline benefits from unprecedented regulatory streamlining. The Nuclear Regulatory Commission finalized its Part 53 risk-informed, technology-inclusive framework on March 25, 2026, effective April 29—five days from now. The new rules, detailed by Perkins Coie, replace decades-old regulations designed for large light-water reactors with pathways tailored to SMR designs.
The NRC completed technical review of X-Energy’s Xe-100 demonstration project at Dow’s Seadrift, Texas facility on an accelerated 18-month timeline, with potential construction permit approval by year-end 2026 according to Utility Dive. This compares to the traditional 10+ year licensing cycle that stalled previous nuclear buildouts.
Administrative pressure reinforced the shift. Executive Order 14300, issued in March 2025, directed the NRC to establish an 18-month deadline for new reactor licensing and 12 months for renewals. NRC data showed 90% of 700 licensing actions in 2025 were completed ahead of schedule—a performance benchmark that would have been unthinkable in the 2010s.
SMRs differ from conventional nuclear plants in scale and modularity. X-Energy’s Xe-100 reactor generates 80 MW—roughly one-tenth the output of a traditional pressurized water reactor—using high-temperature gas cooling and TRISO pebble fuel. The smaller footprint enables factory fabrication, co-location with industrial facilities or data centers, and incremental capacity additions without billion-dollar capital outlays.
Amazon’s Strategic Calculus
Amazon’s $700 million equity investment in X-Energy in 2025, followed by commitments to deploy 5+ GW of new capacity by 2039, positions the company to control dedicated power infrastructure as AI workloads scale. The first commercial X-Energy reactors are slated for the early 2030s, per Energy News Beat.
This aligns with Amazon’s broader data center strategy. The company announced a $20 billion investment in Pennsylvania facilities adjacent to the Susquehanna nuclear plant, effectively locking in long-term baseload power at a time when competitors face rising interconnection queues and utility pushback on grid load.
“Data centers are the critical infrastructure needed to support AI leadership, and their power needs continue to accelerate to meet the growing needs of our customers,” said Vibhu Kaushik, head of worldwide energy at AWS, in a statement reported by Carbon Credits.
Competitive Implications
The nuclear buildout creates asymmetric advantages. Firms that secure dedicated baseload capacity can site data centers without grid dependency, accelerating deployment timelines and reducing exposure to utility rate volatility. Those relying on shared grid resources face lengthening interconnection queues—now averaging 3-5 years in high-demand regions—and potential capacity constraints that could throttle AI model training schedules.
X-Energy’s customer pipeline includes Dow Chemical (Texas demonstration project), Energy Northwest (Washington state), and Talen Energy. The company is also collaborating with Korea Hydro & Nuclear Power and Doosan on a broader deployment initiative, per Carbon Credits. Each partnership represents potential gigawatt-scale capacity that could shift regional power dynamics.
NuScale, another SMR developer, received NRC design approval for its 462 MW US460 reactor in February 2026 after a sub-two-year review—ahead of schedule and explicitly targeting hyperscale data center applications, reported Utility Dive. The competitive landscape is consolidating around firms that can navigate both NRC licensing and tech-sector contracting.
- X-Energy’s 39% pricing premium signals investor confidence in nuclear-AI Infrastructure thesis, not speculative positioning.
- Big tech has contracted 10+ GW of new nuclear capacity in twelve months—more than the entire U.S. solar deployment in 2023.
- NRC regulatory acceleration (Part 53, 18-month licensing) has compressed timelines that previously killed SMR economics.
- Grid saturation in key markets (PJM, ERCOT) makes dedicated power infrastructure a competitive moat, not a commodity input.
- First-mover advantage accrues to firms securing capacity now—interconnection queues and permitting delays will worsen through 2027.
What to Watch
X-Energy shares begin trading on Nasdaq under ticker XE, with the offering expected to close April 27. Early price action will indicate whether public market investors share the IPO’s enterprise valuation or view SMR deployment timelines as too speculative.
The Dow Seadrift project’s NRC construction permit decision—expected by year-end 2026—will serve as the clearest test of whether 18-month licensing timelines are achievable at scale. Delays would signal that regulatory streamlining remains aspirational; on-time approval would validate the investment thesis and likely trigger additional SMR deployments.
Monitor PJM and ERCOT interconnection queue data for colocation trends. If tech firms begin withdrawing renewable-powered data center applications in favor of nuclear-backed projects, it confirms the baseload imperative is overriding cost considerations.
Finally, track whether Microsoft, Meta, or other hyperscalers announce additional nuclear commitments in Q2-Q3 2026. The current 10 GW contracted capacity represents less than 20% of projected 2030 data center demand growth. The real question is not whether nuclear powers AI infrastructure—it’s who controls that infrastructure and at what valuation.