China Has Built What the West Promised

For ten years, industry forums and investor presentations celebrated the small modular reactor as the next major energy breakthrough. Hundreds of projects were announced, billions raised, roadmaps published. Result: no commercial reactor of this type exists yet in terrestrial operation in the Western world. In China, the first one has just been built.

Linglong One, 125 megawatts electric, is set to enter commercial operation in the first half of 2026 on Hainan Island. This will be the first commercial terrestrial small modular reactor in history. Not a demonstration. Not a prototype. A reactor connected to the grid, designed to be replicated.

The Essential Points

  • Linglong One (125 MWe) will reach commercial operation in the first half of 2026, the world’s first commercial terrestrial SMR
  • CNNC (China National Nuclear Corporation) has been developing the ACP100 since 2010; construction began in 2021 in Changjiang, Hainan
  • China is developing at least six other SMR designs in parallel at advanced stages
  • American hyperscalers have signed for more than 10 GW of nuclear power to supply their data centers, with no commercial reactor available in the United States
  • China’s industrial lead on SMRs reproduces the pattern of batteries and solar panels: the first manufacturer to reliabilize production imposes its standards

What “First in the World” Concretely Means

Linglong One did not emerge from nowhere. Development of the ACP100 by China National Nuclear Corporation (CNNC) began as early as 2010, with preliminary design completed in 2014. The official announcement of the construction project came in July 2019, and construction formally began in July 2021 in Changjiang, in Hainan Province. Five years of work for a 125 MWe reactor using pressurized water, proven technology adapted to a compact configuration.

125 megawatts is roughly eight times less than a typical conventional nuclear reactor of 1,000 MWe. It is sufficient to power a mid-sized city, an isolated industrial zone, an island. This is precisely the format’s value: where large reactors require connections to dense grids, decades of planning and capital that few states can mobilize, an SMR can be deployed on the periphery, in poorly connected regions, in countries with limited investment capacity. Modularity is not a fine technical achievement—it is a response to a very concrete geographic and financial constraint.

What “first in the world” means in this context is validation of the concept at industrial scale. A reactor that runs, generates electricity, accumulates operating hours, produces reliability data and supplies the operational feedback essential for building the next one faster and cheaper. This is the leap that neither the United States, nor the United Kingdom, nor France has yet achieved.

Meanwhile, Western Projects Accumulate in Drawers

The gap between announcement and realization has become the defining characteristic of the SMR sector in the West. NuScale Power, the most advanced American company, was still being presented as a global pioneer three years ago. In 2023, its flagship project in Idaho was canceled: estimated costs had more than doubled, potential customers withdrew, and timelines stretched until the end of the decade. Design certification obtained from the Nuclear Regulatory Commission in 2022 was not sufficient to bridge the distance between paper and concrete.

In the United Kingdom, Rolls-Royce SMR is advancing its 470 MWe design and has obtained substantial government commitments. But the first reactors are not expected before 2035 at the earliest, in the best-case schedule scenario. In France, EDF is developing its own concept, NUWARD, but final investment decisions remain suspended on political and regulatory decisions that are dragging on.

This is not a question of technical competence. British, French and American engineers master reactor physics. The problem is structural: Western markets combine fragmented nuclear regulation, high certification costs, industrial supply chains that have lost their skills after decades of disinvestment, and financing structures that do not know how to value an asset with a 40-year lifespan. China has not solved these problems; it has never faced them to this degree.

Demand Is There, Western Supply Is Not Following

The market context has never been more favorable to nuclear power. Major technology companies have transformed their infrastructure model in a matter of years, to the point that data center power supply has become a top-tier strategic constraint. Microsoft, Google, Amazon and Meta have collectively announced commitments for more than 10 gigawatts of nuclear capacity dedicated to their data centers. This figure, compiled by SMR Intel in its “State of SMR 2026” report, represents more than the total nuclear capacity of many European countries.

These companies are not making these announcements out of abstract environmental conviction. Artificial intelligence data centers consume continuous power, day and night, incompatible with the intermittency of solar and wind. Nuclear is the only dispatchable low-carbon source that can meet this load profile at scale. Europe, which knows how to build data centers but struggles to power them, illustrates exactly this bottleneck.

The American problem is straightforward: this 10 GW of signed demand has no corresponding supply available in the short term. Large reactors take twenty years to build in the recent American context. SMRs are still on paper or in early phases of certification. Meanwhile, China accumulates operating hours on the first unit and prepares the next ones.

The Precedent of Batteries and Solar Panels

Recent industrial history provides an analytical framework that Western decision-makers would do well not to ignore. In the 2000s, German, American and Japanese companies dominated solar panel production. China invested massively, absorbed the initial losses necessary for scaling up, standardized its components and reduced costs at a pace no one had anticipated. Result: in 2024, China produces more than 80% of the world’s solar panels, according to International Energy Agency data. The same pattern applies to lithium-ion batteries and electric vehicles.

This is not industrial magic. It is the consequence of a simple fact: the first manufacturer to reliabilize serial production will impose its standards, components, interfaces and maintenance protocols. Customers who buy a first Linglong Two or Three reactor in the 2030s will need Chinese spare parts, technicians trained on Chinese systems, operating procedures documented in Mandarin and then translated. This is how industrial standards are constituted, not by decree.

For nuclear power, the stakes are amplified by asset lifespan. A reactor installed in 2030 will run until 2070. Technological dependence is, by construction, a forty-year dependence. This is not exactly the type of relationship a country concerned with its energy sovereignty would want to maintain with a supplier whose geopolitical relations are unpredictable over that horizon. But the decision to buy is made in the present, with the offers available in the present.

What China Is Doing in Parallel

Linglong One is not an isolated experiment. China is simultaneously developing several other SMR designs at advanced stages. The floating ACP100S reactor, derived from Linglong One, is designed for offshore platforms and remote island applications. Molten salt and high-temperature reactor concepts are under development at the Institute of Plasma Physics of the Chinese Academy of Sciences. CNNC and SPIC (State Power Investment Corporation) each have their own parallel programs.

This diversity is not disorderly. It corresponds to a strategy of covering different market segments: terrestrial networks, offshore zones, exports to developing countries, high-temperature industrial applications for hydrogen production or process heat. Each design tested today is an open option for the next decade.

Export is at the heart of the strategy. China has already exported two large EPR-equivalent reactors to Pakistan via the Chashma agreement and is developing discussions with some twenty countries in Asia, Africa and the Middle East as part of its “Nuclear Silk Road” initiative. SMRs, smaller and financially accessible, considerably expand the pool of potential customers. A country that cannot afford a 1,000 MWe reactor costing $10 billion can consider a 125 MWe SMR at a proportionally lower cost, with lower land footprint and network complexity.

What Remains Open for the West

China’s lead is real. It is not irreversible at all scales, but catching up requires specific conditions that are not yet in place.

The first condition is regulatory. The United States has begun a significant reform of its nuclear certification process. The Nuclear Regulatory Commission is working on accelerated approval pathways for non-light-water designs. The ADVANCE Act, signed in 2024, aims to reduce authorization delays and costs. These reforms are real, but their effects will be measured in years, not quarters.

The second condition is industrial. The United States, the United Kingdom and France still have nuclear engineering expertise, but their manufacturing chains have suffered from decades of construction freeze. Reconstituting the capacity to forge pressure vessel components, manufacture precision steam generators, and train qualified Level 4 welders cannot be decreed. It takes time and sustained investment, not press announcements.

The third condition is financial. Nuclear is a long-lived asset with high construction risk. Private capital markets are structurally reluctant without substantial public guarantees. Countries that are advancing fastest, whether China, South Korea or, to a lesser extent, Canada with its SMR program, are those that have accepted that the state should play an active role in taking initial risk.

A window remains. Markets to be built in the 2030s have not yet chosen their supplier. The data center demand is so massive it could support several players. Countries like Canada, with its BWRX-300 reactor under development in partnership with GE Hitachi, or the United Kingdom with Rolls-Royce, have credible programs. The question is not whether the West can still play in this market. It is how much time it has left to decide to do so seriously.

Linglong One will enter service in a few months. The data it will produce during its first years of operation will be worth more than all the feasibility studies published in the West since 2010. The race for standards has begun. It has a leader.

Sources

  1. SMR Intel, State of SMR 2026 : https://smrintel.com/state-of-smr-2026/
  2. Wikipedia, Linglong One : https://en.wikipedia.org/wiki/Linglong_One
  3. International Energy Agency (IEA), Solar PV Global Supply Chains, 2022
  4. Nuclear Regulatory Commission (NRC), ADVANCE Act Implementation, 2024
  5. World Nuclear Association, Nuclear Power in China, updated 2025
  6. CNNC official — construction start Linglong One (July 2021) : https://en.cnnc.com.cn/2021-07/14/c_1023245.htm
  7. CNNC official — cold functional testing completed (October 2025) : https://en.cnnc.com.cn/2025-10/24/c_1134890.htm
  8. World Nuclear News — cold testing Chinese SMR : https://www.world-nuclear-news.org/articles/cold-testing-of-chinese-smr-completed
  9. U.S. Department of Energy — NRC certification NuScale 2022 : https://www.energy.gov/ne/articles/nrc-certifies-first-us-small-modular-reactor-design
  10. Wikipedia — NuScale Power — Idaho cancellation and costs : https://en.wikipedia.org/wiki/NuScale_Power
  11. Statista / IEA PVPS — China’s share of global PV production 2024 : https://www.statista.com/statistics/668749/regional-distribution-of-solar-pv-module-manufacturing/
  12. IEA — China’s share in global solar manufacturing : https://www.iea.org/reports/solar-pv-global-supply-chains/executive-summary
  13. SMR Intel — hyperscaler nuclear deals tracker (May 2026) : https://smrintel.com/nuclear-data-center-deals/
  14. Carnegie Endowment — assessment of hyperscaler nuclear commitments (June 2026) : https://carnegieendowment.org/research/2026/06/beyond-the-hype-assessing-hyperscaler-nuclear-commitments-against-us-energy-realities
  15. Al Arabiya / Reuters — commercial operation startup Linglong One H1 2026 : https://english.alarabiya.net/News/world/2025/12/11/china-to-begin-commercial-operation-of-first-small-modular-nuclear-reactor-in-2026