Countdown to nuclear fusion power generation: The national team has only scheduled it for 2045. Why do private enterprises dare to claim it will be achieved by 2033?

Private capital is flowing into the nuclear fusion track. The commercialization timeline of private companies is 12 years ahead of the national team, sparking debates about acceleration and bubbles.

In January 2026, a financing of 1 billion yuan was transferred to the account of StellarRing Fusion. This company, which has been established for less than five years, has just refreshed the record for a single - round financing in private nuclear fusion in China. The leading investor is the Future Industry Fund under Shanghai State - owned Investment Corporation. According to their plan, a commercial demonstration reactor will be built around 2033, which is a full 12 years ahead of the timeline announced by Duan Xuru, the chief scientist of CNNC, for the national team.

This time difference is becoming the most delicate tension in the nuclear fusion track. The year 2025 is regarded as the first year of fusion commercialization in the industry. The scale of private fusion financing in China has ranked second globally. The total public financing in the first half of 2025 exceeded 11.5 billion yuan, while this figure was almost zero before 2019. A number of enterprises such as ENN Science & Technology, Energy Singularity, StellarRing Fusion, and Nova Fusion have risen rapidly, deploying diverse technological routes from high - temperature superconducting tokamaks to field - reversed configurations and magnetic compression synergy.

Duan Xuru, a deputy to the Two Sessions, presented the official timeline at the Two Sessions: the combustion experiment will start in 2027, the engineering experimental reactor will be built in 2035, and the commercial demonstration reactor will be realized in 2045. However, the capital market obviously doesn't plan to follow this rhythm. The question is: Does the influx of private capital mean industry acceleration, or could it turn into another bubble?

What are hundreds of billions of capital betting on?

Going back to 2017, when ENN Group first started investing in fusion research, there was hardly any other private enterprise in China daring to enter this field. Eight years later, ENN has cumulatively invested 4.5 billion yuan in R & D and built the "Xuanlong - 50U" experimental device. In 2025, it achieved the world's first high - confinement mode discharge of hydrogen - boron plasma. This clean - energy enterprise originating from Hebei has chosen the hydrogen - boron fusion route, which is different from the mainstream deuterium - tritium fusion, and is claimed to be "clean, safe, with easily available fuel and low cost".

StellarRing Fusion, founded in 2021, takes a different path. Its founding team is from the Department of Engineering Physics of Tsinghua University and has chosen the miniaturized spherical tokamak route. Industry calculations show that the cost of building a device with an energy gain Q > 1 for traditional large - scale tokamaks exceeds 15 billion yuan, while the goal of the miniaturized route is to reduce the cost of similar devices by an order of magnitude. This cost difference is the core logic for capital to place bets.

The speed of capital influx is indeed astonishing. The annual financing amount exceeded 5 billion yuan in 2023, and in the first half of 2025, it exceeded 11.5 billion yuan. However, Chen Zhongyong, one of the first batch of experts in the ITER program of the Ministry of Science and Technology, gave a sober judgment: For most of the technological routes where this money is flowing, the scientific feasibility has not been fully verified, let alone engineering implementation.

Where does the time difference between the national team and private enterprises come from?

Duan Xuru divides the commercialization of nuclear fusion into six stages: principle exploration, large - scale experiment, combustion experiment, experimental reactor, demonstration reactor, and commercial reactor. China is currently in the third stage, the "combustion experiment stage". The China HL - 2M Tokamak has just achieved a "double - 100 - million - degree" operation with a nuclear temperature of 117 million degrees Celsius and an electron temperature of 160 million degrees Celsius. According to his deduction, each of the next three steps will take about a decade of development.

However, private enterprises obviously don't have this patience. StellarRing Fusion's plan is: the NTST device will start construction in Shanghai in 2026, complete engineering verification in 2028, and a commercial demonstration reactor will be built in 2033. This rhythm is a full round faster than that of the national team.

The core of the gap lies in the different definitions of "commercialization". Duan Xuru's standard is large - scale power supply at the grid level, which requires considering a series of issues such as the maturity of the industrial chain, economic affordability, and regulatory adaptability. Private enterprises may be targeting more flexible distributed scenarios - Nova Fusion directly aims at AI data centers, with a single - reactor output of 50 - 100 MW and a target cost per kilowatt - hour controlled below 0.1 yuan. This cost line means that fusion power generation should have the ability to compete directly with thermal power, wind power, and photovoltaics.

The collision of these two logics is reshaping the industry landscape. Yan Jianwen, a member of the National Committee of the Chinese People's Political Consultative Conference and the chairman of Fusion New Energy, observed that there is still room for optimization in the performance of existing low - temperature superconducting materials, and high - temperature superconducting materials have become a key area for breakthrough. High - temperature superconducting magnets are precisely one of the core technologies that private enterprises are betting on. Duan Xuru also admitted that if there is significant progress in this area, it is expected to make the fusion reactor more compact and shorten the cycle.

Who will foot the bill when the industrial chain is not yet formed?

In March 2026, a conference on the future industry of controllable nuclear fusion was held in Changzhou Economic Development Zone. Jinchuang Group, an enterprise engaged in rail transit equipment, signed an agreement on site with the Institute of Fusion and Plasma Research of Huazhong University of Science and Technology to establish a joint - venture company to develop a plasma disruption prediction system.

This scenario is quite interesting - a manufacturing enterprise that has little to do with nuclear fusion is starting to get involved. The underlying logic is: If you don't secure a position now, there will be no chance when the industrial chain is formed.

Similar situations are playing out in multiple cities in the Yangtze River Delta. Shanghai has become a gathering place for fusion enterprises, with five fusion companies setting up in 2025 alone. Xu Guosheng, the deputy director of the Institute of Plasma Physics of the Chinese Academy of Sciences, analyzed that different regions are developing in a complementary way based on their own advantages - Shanghai focuses on finance, Hefei on R & D, and Changzhou on manufacturing.

However, the problem is that the industrial chain has not really taken shape. Xu Guosheng said bluntly that driven by the construction of large - scale engineering devices, the upstream and downstream of the industrial chain have begun to develop initially, but "a complete and mature industrial chain has not been formed, and positive economic feedback has not been achieved". A person from a fusion enterprise revealed that as downstream purchasers, they had to personally carry out work that should have been done by the supply chain, such as "manually making" magnets.

This is the most realistic challenge behind the private - enterprise boom. Capital can flow in quickly, and technological routes can blossom in multiple directions, but nuclear fusion is ultimately a project that requires a huge industrial chain to support. Problems such as plasma disruption, materials resistant to high - energy neutron bombardment, and tritium circulation systems - these tough nuts will not disappear automatically just because the financing amount increases.

Yan Jianwen listed a series of problems at this year's Two Sessions: The performance of superconducting materials needs to be improved, the reliability and stability of materials need to be strengthened, there is a prominent shortage of professional talents, and regulations related to tritium management need to be clarified urgently. The Atomic Energy Law came into effect in January this year, but the supporting standards are far behind the pace of industrial development.

Is rapid commercialization really possible?

The integration of AI technology is changing R & D efficiency. Duan Xuru mentioned that artificial intelligence has been initially verified in plasma operation monitoring, control, and instability prediction, and is expected to solve the plasma control problem. Some practitioners said that AI can currently improve efficiency by 20 - 30%, but the expectation is to double or triple it.

Another variable is the catalysis of the industrial chain brought about by the entry of the national team. Since 2023, two national fusion teams have been established - Fusion New Energy, led by the Institute of Plasma Physics of the Chinese Academy of Sciences, has a registered capital of 14.5 billion yuan, and China Fusion Energy Co., Ltd., with the Southwestern Institute of Physics of CNNC as the foundation, has a registered capital of 15 billion yuan. Some industry insiders observed that "now many suppliers are willing to enter this industry even if they make losses. This was rarely seen before."

However, the obstacles to rapid commercialization remain formidable. A report from the International Atomic Energy Agency shows that nearly 40 countries around the world are promoting fusion programs, but the commercialization of fusion energy still faces multiple challenges. The cost of the ITER project has soared from the initial $5 billion to over $22 billion, and the timeline has been repeatedly postponed, sounding an alarm for all aggressive commercialization timelines.

In the view of industry insiders, the successful path of China in the fields of wind power, photovoltaics, and new - energy vehicles may be replicated - before the commercialization of a new technology, guide industrial capabilities to the industrial chain through policies, and unleash strong market competitiveness. This is also the reason why many private enterprises are confident about their goals around 2030.

However, in Xu Guosheng's view, to transform from a major fusion research country to a major fusion industry country, it is necessary to form commercially viable and replicable technological solutions in key areas. And this is precisely not a problem that can be solved by the scale of financing.

In 2026, a double - game is playing out in the nuclear fusion track. Private capital is betting on the speed at which the technology curve is bent by AI and high - temperature superconductivity, while the national team is adhering to the engineering laws precipitated by six decades of scientific research. The timelines of the two sides differ by 12 years, but neither side dares to easily say that the other is wrong.

After all, in the matter of the "artificial sun", humanity has been proven wrong too many times by the "perpetual 50 - year" prediction.

This article is from the WeChat official account "Foreseeing Energy". Author: Foreseeing Energy. Republished by 36Kr with permission.