It’s been a wild ride for anyone following the small modular reactor (SMR) space. If you've been tracking NuScale Power news 2025, you know it’s been a year defined by massive regulatory wins, a few "whoa" moments on the stock market, and a growing obsession with how we're going to keep AI data centers from crashing the power grid. Honestly, it feels like the company is constantly trying to outrun its own reputation. One day they're the darlings of the "Nuclear Renaissance," and the next, short-sellers are breathing down their necks because they haven't poured concrete on a commercial site yet.
But here's the thing. While the "paper reactor" jokes still fly around on Twitter, the reality on the ground in 2025 has shifted. We aren't just talking about PowerPoints anymore.
The Big 77 MWe Win and the Regulatory Hurdle
Most people forget that the original NuScale design was for a 50-megawatt module. That was the one that got the first-ever NRC certification. But 50 megawatts isn't exactly a powerhouse when you're trying to convince a massive utility to switch from coal.
Back in May 2025, the U.S. Nuclear Regulatory Commission (NRC) dropped a bombshell—in a good way. They officially approved the uprated 77 MWe design. This was huge. It basically means the exact same physical footprint can now crank out way more electricity. Specifically, a six-module plant (which they call the VOYGR-6) can now hit about 462 MWe.
Why does this matter for NuScale Power news 2025? Because it fixes the math. The previous 50 MWe design struggled with "economies of scale." By bumping the output by over 50% without a massive redesign, the cost-per-kilowatt-hour finally starts looking like something a CFO wouldn't laugh at.
Data Centers and the 6-Gigawatt Dream
You can’t talk about nuclear in 2025 without talking about AI. Big Tech is desperate. They need 24/7 "baseload" power, and wind or solar—bless them—just don't cut it when the sun goes down and ChatGPT users are still hallucinating recipes.
The standout news from late 2025 was the landmark agreement between NuScale’s partner, ENTRA1 Energy, and the Tennessee Valley Authority (TVA). We are talking about a program to deploy up to 6 gigawatts of SMR capacity.
"This isn't just a pilot project," a tech analyst recently told me. "It's a blueprint for the re-industrialization of the U.S. south."
It’s an ambitious play. The plan is to use these modular reactors to power AI hubs and semiconductor plants. But don't expect these to pop up overnight. The timeline for the first Inter One energy plant delivering to TVA is still pegged for 2030. That’s the "nuclear reality check" nobody likes to hear: physics is fast, but permitting is slow.
The Romania Factor: RoPower and Doicești
While everyone is looking at Idaho or Tennessee, the real action might actually be happening in Romania. NuScale has been heads-down on the RoPower project in Doicești.
Throughout 2025, Fluor (NuScale's majority owner and EPC partner) has been grinding through the Phase 2 Front-End Engineering and Design (FEED) study. They’re basically turning a former coal plant into a high-tech nuclear site.
- Timeline: FEED Phase 2 is slated for completion by the end of 2025.
- Funding: The US EXIM Bank and DFC have already dangled billions in potential support.
- Jobs: We're looking at roughly 2,000 construction jobs once they actually break ground.
If Romania reaches a Final Investment Decision (FID) in early 2026, it will be the first time a NuScale plant actually feels "real" to the global market. It’s the lead domino. If it falls, Poland and the Czech Republic are likely next.
The Financial Rollercoaster: Cash vs. Confidence
Now, let’s talk about the elephant in the room—the stock ($SMR). 2025 was a year of extreme highs and stomach-churning lows.
In late 2025, NuScale reported a massive earnings miss that sent the stock sliding nearly 15% in a single day. Their EPS was way off. Why? Well, they had to recognize a massive $495 million milestone payment to ENTRA1. Basically, as they move closer to commercialization, the accounting gets messy.
But if you look at the balance sheet, they ended the third quarter of 2025 with over $753 million in cash. They aren't going broke anytime soon. They’ve been using an "At-The-Market" (ATM) program to sell shares and stack cash while the hype is high. It’s smart, even if it dilutes the current shareholders a bit.
✨ Don't miss: Cleaning Your Litter-Robot 4: What Most People Get Wrong About the Deep Clean
Then you have Fluor. In November 2025, Fluor announced they were going to monetize their remaining 39% stake in NuScale by mid-2026. Some investors panicked, thinking Fluor was "abandoning ship." But honestly? Fluor is an engineering firm, not a venture capital fund. They’ve made billions on their initial investment and want to buy back their own stock. They’re still the lead contractor for the projects. It’s a transition from "parent company" to "commercial partner."
What Most People Get Wrong About NuScale
The loudest critics always point to the canceled Utah project (CFPP) from a few years ago. "They couldn't build it there, so they can't build it anywhere," goes the logic.
That's a bit reductive. The Utah project failed because interest rates spiked and the price of steel went through the roof during a very specific inflationary window. NuScale 2.0 (the 2025 version) is leaner. They’ve moved away from trying to be the "owner/operator" and are now more like a software company that licenses the "operating system" (the reactor design) to developers like ENTRA1.
Current Project Status Breakdown
| Project Site | Status as of late 2025 | Target Operation |
|---|---|---|
| Doicești, Romania | FEED Phase 2 nearly complete | 2029-2030 |
| TVA (USA) | 6GW Framework Agreement signed | 2030+ |
| Standard Power | Data center partnership active | Ongoing design |
| Varduhi, Armenia | Strategic MOU stage | TBD |
Moving Toward 2026: What's Next?
If you’re tracking NuScale Power news 2025, you shouldn't just be looking at the stock price. You need to look at the supply chain.
NuScale has been working with Doosan Enerbility in South Korea to actually forge the long-lead materials for the reactors. You can’t just 3D print a nuclear pressure vessel. These things take years to bake. The fact that Doosan is already spinning up production is a much better indicator of "reality" than any press release.
As we head into 2026, keep your eyes on the Romanian FID. That is the only thing that will truly silence the skeptics. Until then, expect more volatility and more debates about whether SMRs are the future or just a very expensive dream.
Actionable Insights for Following NuScale:
- Watch the NRC Registry: Now that the 77 MWe design is approved, look for "Combined License" (COL) applications from specific sites. That’s when the clock really starts.
- Monitor Fluor’s Exit: The sell-off of Fluor’s 111 million shares will likely keep a lid on the stock price through Q2 2026. If the price stabilizes after that, it's a sign of institutional strength.
- Track Data Center PPAs: The first time a "Hyperscaler" (think Google, Amazon, or Microsoft) signs a direct Power Purchase Agreement for a NuScale-powered plant, the game changes forever.
- Ignore the "Paper" Labels: Focus on the "long-lead material" orders. If NuScale is paying for forged steel, they're planning to build.
Nuclear is a slow-motion industry. NuScale is currently the only one with a cleared path through the American regulatory woods, but being first also means they're the ones hitting all the cobwebs. 2025 proved they have the stamina; 2026 will prove if they have the market.