The spread of artificial intelligence and rising power needs of data centers are accelerating interest in small modular reactors (SMRs). More governments and major companies are backing SMR projects, but questions remain over project economics, financing risks and secure fuel supplies.
Projects and key numbers
In Canada, one of the first four SMRs is under construction at Darlington, Ontario. The overall project is estimated to cost $15 billion, and planners expect to connect the first unit to the grid by the end of the decade. A single reactor could provide electricity for as many as 300,000 households.
Investment flows reflect growing interest: PitchBook reported a record $2.9 billion of private capital into the sector last year. The U.S. government has pledged about $3 billion over the past year for development and fuel security; the European Commission considers SMRs an important future element of the EU energy mix and plans deployments from the 2030s. The U.K. is supporting Rolls-Royce’s project in Wales.
Cost and financing risks
Economic viability is still uncertain. Mark Winfield, a professor at York University, told the Financial Times that costs for a first-of-a-kind project are hard to pin down, and taxpayers and electricity consumers may ultimately bear cost overruns. During construction, Ontario Power Generation can incorporate financing costs into electricity prices before production starts; the company aims to recover nearly $800 million in this way between 2027 and 2031.
Limited real-world experience increases risk: X-Energy’s demonstration project was estimated at $2.5 billion in 2021 but rose to nearly $6 billion by 2023. NuScale cancelled its initial U.S. project amid rising costs and insufficient demand, despite $183 million of federal support, and a supervisory agency found that critical risks had not been properly assessed.
Fuel supply and geopolitical constraints
Fuel availability is another potential bottleneck. Several developers, including Oklo and X-Energy, plan to use higher-assay low-enriched uranium (HALEU). Commercial HALEU production outside China and Russia is currently limited.
The U.S. is investing $2.7 billion to build domestic enrichment capacity, and Urenco plans its first European HALEU facility by 2031. By contrast, the Darlington project uses GE Vernova–Hitachi’s evolved boiling-water reactor technology and intends to rely on conventional nuclear fuel, reducing dependence on new fuel supply chains that many SMR designs would require.
Regional developments and Hungary
Rolls-Royce has identified Hungary as a priority market for its SMR and has been in discussions with the Hungarian government and energy companies since 2022. The company has contacted more than 40 Hungarian suppliers, and regional reactor programmes are taking shape. Rolls-Royce’s 470 MW SMR units could play a role in replacing conventional nuclear capacity.
Hungary’s Országos Atomenergia Hivatal sent a delegation to the U.S. nuclear regulator to learn about SMR regulation; Hungarian authorities state they would deploy only proven technologies and do not intend to become a testing ground.
Outlook and comparative costs
The International Energy Agency (IEA) estimates that by 2050 SMRs could add between 40 and 200 gigawatts of new capacity worldwide, depending on the pace of decarbonisation. Some developers’ cost estimates remain high: Oklo recently suggested levelized costs of electricity of $80–$130/MWh for its design, compared with average utility-scale solar costs of about $45/MWh in 2025.
Conclusion
Growing electricity demand from AI and data centers, together with rising investor interest, is pushing SMRs into the spotlight. Nevertheless, significant hurdles—uncertain construction costs, financing arrangements that can shift risk to public consumers, and limited HALEU supply chains—remain barriers to broad deployment.
