01NuScale Power · NYSE: SMR
The holy grail of AI-era energy, or a capital bubble?
A company that has never built a reactor is being treated as the answer to the AI power gap. Scarcity, tailwind and its own record of failure, side by side.
Timestamp. This note was written on 29 November 2025. Share price, cash, backlog and the competitive picture are as of that date. Nothing was updated when it was moved here; read it as a dated judgement.
Disclaimer · This material is for personal research and educational discussion only. It is compiled from public information and does not constitute investment advice, an offer, or a recommendation, nor a basis for buying or selling any security. The presenter may hold positions in the securities mentioned. Investing involves risk; make your own judgement and bear the consequences of your own decisions.
02The view in one page
A high-risk hunt for a "unicorn"
Rating
Speculative Buy
Small position only
It is the "early Tesla" of nuclear: ahead on technology, but its life still depends on ramping production and delivering.
- Scarcity: the only SMR design in the world with full US NRC certification.
- Tailwind: AI data centres urgently need round-the-clock clean power, and SMRs are one of the best answers.
- Risk: early commercialisation, no profits, and a flagship project (UAMPS / CFPP) that already failed once.
03The business
What does NuScale actually do?
Core product · the NuScale Power Module (NPM)
- What it is: a small modular reactor (SMR).
- How it's built: factory-made, shipped by truck, assembled on site.
- Safety: passive systems that cool the reactor without electricity, avoiding a Fukushima-style failure.
Business model · asset-light
- Does not build or own power plants.
- Earns from technology licensing, module sales and engineering services.
- The triangle: NuScale (technology) + Doosan (manufacturing) + Fluor (engineering and construction).
04The AI energy crisis
Why SMRs, and why now?
Demand
AI compute is exploding; a ChatGPT query uses roughly ten times the electricity of a Google search. Data centres need stable power 24/7.
Supply
Wind and solar depend on the weather; conventional large nuclear takes a decade to build.
The role of SMRs. Baseload power that complements wind and solar, and small enough to be built on the site of a retired coal plant, reusing the existing grid connection.
05Moat and catalysts
Regulation is the moat; orders are the catalyst
Regulatory moat
- NRC certification: the first company in history to receive Standard Design Approval.
- Lead: three to five years ahead of X-energy, Oklo and others on the regulatory track.
Pipeline
- TVA (Tennessee Valley Authority): a major cooperation agreement with gigawatt-scale potential.
- RoPower (Romania): Europe's first SMR project, backed by US government funding.
- Standard Power: a 2 GW plan dedicated to data-centre power.
06Key risks and the record
The technology passed. The economics haven't yet.
The Utah project (CFPP) failed
Cancelled in late 2023 after costs jumped from $58/MWh to $89/MWh. The lesson: even with the technology approved, being too expensive is still the biggest obstacle.
Slides, no product
No module has been built and operated. All revenue is still "engineering services".
Dilution
The company keeps burning cash, with a heavy Q3 loss, and will very likely raise equity again, diluting existing holders.
07Financials
How much longer does the cash last?
Cash
~$690M
As of the latest report
Long-term debt
0
No debt burden today
Quarterly revenue
$10–50M
Lumpy
Quarterly net loss
$40–50M
Roughly three to four years at this rate
How to value it. Not on a P/E. Look at EV / backlog, enterprise value over the order book. The market is paying for expectations beyond 2030.
08Competition
Who is catching up?
| Competitor | Technology | Status | Strengths / weaknesses |
|---|---|---|---|
| NuScale (SMR) | Light-water reactor | NRC approved | Furthest along, but higher cost |
| GE Hitachi | BWRX-300 | In design | Backed by giants, strong supply chain, Canadian orders in hand |
| Oklo (OKLO) | Fast reactor | Early | Backed by Sam Altman, more radical technology, slower regulatory path |
| X-energy | High-temperature gas-cooled | Early | Strong industrial use case, partnered with Dow Chemical |
09Watch list
What to watch next
Short term · 6 to 12 months
- Whether the Romanian project reaches the next stage (FID, final investment decision).
- Whether a large technology company announces it will adopt NuScale's technology directly.
Medium term · 1 to 3 years
- Manufacturing of the first module starting at the Doosan plant.
- Whether the timeline to positive cash flow moves forward.
10Final thoughts
At the intersection of the nuclear revival and AI power demand
It is the purest and most advanced SMR name on the US market. But what it sells lives beyond 2030, and every step between here and there can go wrong.
If you believe AI needs nuclear and can stomach 50% swings
A small position, say 2 to 5%.
If you want steady dividends
Stay away.
Written 29 November 2025. Personal research, not investment advice. Investing involves risk; make your own decisions.