The market cap peaked above $10 billion in 2025. The company has not sold a single commercial reactor. This is the defining anomaly of NuScale Power's post-SPAC life.
NuScale went public via a SPAC merger in May 2022, raising roughly $380 million. At the time, the company was the first and only SMR developer to receive U.S. NRC design certification. That distinction, however, has not translated into commercial traction. The flagship Carbon Free Power Project in Idaho was terminated in November 2023 after costs ballooned. The LCOE estimate jumped from $58/MWh to $89/MWh in a single year, and the utility partners walked away. NuScale was left with a certified design and zero paying customers.
Fast forward to 2026. The stock trades at a level that implies the market believes NuScale will dominate the coming AI data center power boom. The narrative is seductive: AI accelerates SMR design, AI data centers need 24/7 carbon-free power, and NuScale is the public-market proxy for this convergence. The logic is a closed loop. The problem is that the loop describes a future that is at least a decade away, while the market is pricing it as if it were imminent.
My assessment is based on a forensic review of the SMR supply chain, licensing timelines, and competitive positioning. I have watched this industry cycle through hype phases before. The 2017 ICO boom taught me that a compelling narrative without a working product is a liability, not an asset. The same discipline applies here.
The core thesis is straightforward: AI can accelerate the design validation phase, but it cannot accelerate the licensing timeline, the supply chain build-out, or the first-of-a-kind construction learning curve.
Let me break down the specifics.
The AI Design Accelerant Is Real, But Marginal
NuScale has positioned 'AI-accelerated design' as a headline feature. The technical reality is more measured. AI applications in nuclear engineering fall into four tiers: parameter optimization, simulation acceleration, safety analysis, and documentation generation. The first and last tiers are mature. The middle two are promising but unproven under regulatory scrutiny.
The video above shows Kuok Khoon Hong, Chairman and CEO of Wilmar International, discussing the company's strategy at the Credit Suisse Asian Investment Conference in Hong Kong, March 26, 2013. Wilmar is one of Asia's leading agribusiness groups.
The regulatory reality is that the NRC requires traceable physical verification for any design change. An AI model that suggests a new fuel arrangement must be backed by experimental data. The NRC's licensing process is not designed to accept black-box outputs. A divorce can absorb 20-40% of engineering time, but it does not compress the critical path of regulatory review.
The real bottleneck was never design speed. NuScale received its final Safety Evaluation Report in January 2023, after a seven-year review process. The next SMR applicant will not have to wait that long—the NRC has streamlined the process—but it will still take years, not months. Design certification is only the first step. Construction permits, operating licenses, and site-specific evaluations all remain.
The supply chain is a harder constraint. High-Assay Low-Enriched Uranium (HALEU) is the fuel required for most advanced reactors. Current U.S. production capacity is minimal. The DOE has funded Centrus Energy to build a demonstration cascade in Ohio, but meaningful commercial capacity is not expected until 2027-2028. Russia is a major HALEU supplier, which creates a geopolitical contradiction that cannot be solved by AI.
The manufacturing ecosystem is equally immature. Reactor pressure vessels, steam generators, and control rod drive mechanisms are manufactured by a handful of specialized firms in the U.S., Japan, South Korea, and Europe. These supply chains were built for large-scale custom reactors, not standardized modular production. Retooling for volume manufacturing requires billions in capital investment. No company will make that investment without a firm order book. NuScale's current MOUs with Romania, Poland, and Kazakhstan have not reached final investment decisions. The company's cash runway is approximately 1.5-2 years at current burn rates (about $250 million annually against a few tens of millions in revenue).
The competitive landscape intensifies the pressure. China's ACP100, the world's first land-based commercial SMR, is scheduled for grid connection in 2026. Russia's KLT-40S floating units have been operating since 2020. Both are products of state-backed enterprises that do not need to demonstrate quarterly shareholder returns. NuScale's SPAC structure and 'AI + nuclear' market positioning make it a high-beta thematic stock, not a utility-grade infrastructure play.
The contrarian angle is uncomfortable but necessary: the market may be overvaluing the U.S. SMR champion based on narrative alignment with the AI boom, while undervaluing the actual industrial progress of state-owned competitors. The U.S. policy framework is the most supportive in history—tax credits under the IRA, DOE ARDP funding, and accelerated licensing pathways—but policy enthusiasm does not equal commercial execution.
What are the concrete catalysts to watch? First, any announcement of a signed power purchase agreement with a major hyperscaler that includes a specific deployment timeline and site selection. Second, progress on the U.S. HALEU supply chain. Third, evidence that NuScale's new design iteration (NPM-20) has materially reduced projected construction costs below the $89/MWh threshold that killed the Idaho project. Fourth, the successful commercial operation of the Chinese ACP100, which will provide the first real-world data on SMR economics.
The time mismatch is the fundamental risk. Hyperscalers need new power capacity now. SMR deployment at scale will not occur before 2030. Between now and then, natural gas with carbon capture and expanded grid interconnections will fill the gap. NuScale's addressable market may not materialize until the early 2030s, assuming no major technical or regulatory setbacks. The current share price impatience is not a strategy.
I audit the code, not the charisma.
Yields are calculated, not guaranteed.
Diversification is the only safety net.
Strategy beats speculation every time.
The next 18 months will determine whether NuScale is a technology leader with a timing problem or a narrative stock with no exit. Watch the order book, not the headlines. The data, when it arrives, will tell you everything.