The AI computing power arms race is increasingly highlighting energy supply as a critical issue. The nuclear fusion sector has seen a wave of investment in recent years, with companies like StarPower Fusion and Energy Singularity setting new funding records and expanding the potential for capital in the "ultimate energy source." Now, a different path is emerging: nuclear fission Small Modular Reactors (SMRs).
On July 23, 2026, Shanghai Junhe Atomic Technology Co., Ltd. announced the completion of several hundred million yuan in Series Angel funding. This marks the most significant Angel round in the domestic nuclear fission SMR sector to date. The funding was led by Sequoia Capital China, with participation from state-owned capital and financial institutions such as the Shanghai Future Industry Fund, Shanghai Disruptive Technology Fund, Lingang Sci-Tech, and Guotai Haitong. Industrial capital, like the Vision Carbon Neutrality Fund, also joined, alongside market-driven hard-tech investors including Zhongke Chuangxing, Houshi Capital, Dongfang Fuhai, Juhe Capital, Delian Capital, Jinshajiang United, Qiyu Yulin, Dawu Venture Capital, Yuntai Capital, and Guangyue Investment. Taihe Capital served as the financial advisor for subsequent rounds.
This is the third funding round for Junhe Atomic within a year of its founding. The initial Angel round in October 2025 was exclusively invested by Zhongke Chuangxing, providing early-stage incubation for the technology and team. On May 25, 2026, the company successfully completed another Angel round of several hundred million yuan, led by a leading market-oriented institution, with 14 entities including Lingang Sci-Tech, Houshi Capital, Dongfang Fuhai, Juhe Capital, Delian Capital, Guotai Haitong, Jinshajiang United Capital, Qiyu Yulin, Dawu Venture Capital, the Sequoia Vision Carbon Neutrality Fund, Yuntai Capital, Disruptive Fund, and Guangyue Investment. Completing three rounds of funding, with amounts reaching hundreds of millions of yuan and dozens of institutions investing, in just nine months for a hard-tech nuclear energy early-stage project is rare. This underscores the long-term market confidence in the SMR and micro-reactor track.
Taihe Capital stated, "This is not just a funding round; it reflects the primary market's new judgment on energy infrastructure for the AI era. Nuclear fusion represents the future, while nuclear fission SMRs are addressing the practical needs of the next decade."
A 70-Year-Old Nuclear Veteran's Second Venture
The nuclear energy industry is a hard-core track heavy on qualifications, engineering, and experience. Unlike fast-iterating sectors like internet and AI, the core barrier to entry for nuclear power startups is not just single-point technological innovation, but complete engineering experience, compliance review capabilities, supply chain resources, and project execution ability. Junhe Atomic's ability to secure heavy investment from top capital immediately upon founding is primarily due to its star, established founding team, entirely avoiding the common pitfall of "having technology but no engineering, having R&D but no deployment."
Founder Tian Jiashu is a significant name in China's nuclear industry. A State Council Special Allowance expert, he has served as Director of the Ministry of Ecology and Environment's Nuclear and Radiation Safety Center, Deputy Chief Engineer of CNNC (China National Nuclear Corporation) and CGN (China General Nuclear Power Group), and General Manager of Hualong International. He is also a nuclear safety advisor to the International Atomic Energy Agency. He was the core leader in the localization and overall design of the "Hualong One" nuclear technology, a national symbol for China's nuclear power industry going global. He has experienced the complete chain of a nuclear power plant from design to construction and commercial operation.
In 2025, the retired Tian Jiashu chose to start a new venture, founding Junhe Atomic in Shanghai. His motivation: as energy demand becomes more decentralized and intense outside of large nuclear power plants, can nuclear energy transform from a large-scale national project into a more standardized and flexible energy product? This idea is backed by an ongoing structural shift. The "power black hole" of AI data centers, the low-carbon transition of industrial parks, and the rigid demand for continuous power supply in high-end manufacturing all require not a large nuclear plant with billions in investment and a decade-long construction period, but a modular energy product that can be flexibly deployed, expanded in phases, and located close to the user.
Soon, Tian Jiashu assembled a high-caliber team, including seasoned professionals from CGN, State Power Investment Corporation, Huaneng Group, and EDF (Électricité de France), with an average industry experience of over 20 years. Investors describe this as an "All-Star" team capable of "design, construction, review, operation, and project development." Within a year of its founding, Junhe Atomic's R&D team covers key areas including reactor physics, thermal hydraulics, safety analysis, nuclear island systems, instrumentation and control, equipment manufacturing, engineering construction, and project development. For advanced nuclear energy startups, the most scarce resource is not people who can design a reactor, but those who know how to build, manage, and operate it.
SMRs, by international definition, are reactors with a single-unit electrical power not exceeding 300 MWe, designed for modular construction. Compared to traditional large nuclear plants, SMRs can first achieve a standardized design, then deploy one or more modules based on customer load, expanding capacity gradually, thus lowering the construction barrier. Among the global major SMR technology routes, Junhe Atomic has chosen to invest in the advanced small pressurized water reactor (PWR) technology path.
The Biggest Risk
Most advanced reactor designs require HALEU (High-Assay Low-Enriched Uranium, enrichment 5%-20%) fuel. The HALEU supply gap is the single biggest risk for the advanced nuclear industry. The US currently produces only about 900 kg per year, aiming for 6 metric tons/year within 42 months. Russia was the only commercial HALEU supplier, and after the US banned Russian enriched uranium imports in 2024, domestic production became urgent. This means SMRs using light water/low enriched uranium fuel, like the Junhe-1, have a significant advantage in fuel supply chain reliability.
This is because the pressurized water reactor is one of the most widely used commercial nuclear power technologies globally, with mature design experience, fuel systems, equipment supply chains, and safety regulations, resulting in low engineering risk. For a startup, a mature technology path can significantly reduce R&D trial-and-error costs and shorten the time to deployment, allowing the company to focus its core resources on modular integration, safety system optimization, miniaturization adaptation, and customized transformation for specific scenarios, rather than validating fundamental principles.
On this basis, Junhe Atomic has achieved two core technological innovations to upgrade the traditional PWR. The first is its passive natural circulation technology, a core technical barrier for the Junhe-1. Traditional PWRs rely on primary pumps to drive water circulation, resulting in complex equipment, high maintenance costs, and high risk of failure. By optimizing the structure and upgrading the thermal-hydraulic system, the company has achieved pump-free natural circulation, relying on natural fluid heat transfer for energy transmission, greatly simplifying the system, reducing equipment costs and maintenance difficulty, while improving reactor stability and intrinsic safety. This is a key differentiator for small PWRs.
The second is the integrated pool-type design with a fully passive safety system. By integrating the core components of the pressure vessel, pressurizer, and steam generator, the reactor volume is reduced, making it suitable for modular prefabrication. The fully passive safety system requires no human intervention or external power, autonomously providing safety protection under extreme conditions, thoroughly addressing the safety concerns of small reactors and meeting the requirements for deployment near urban areas or industrial parks. The Junhe-1 (100 MWe small PWR) has completed its full conceptual design, and the company is advancing optimization of the core physics, thermal hydraulics, and main system, while developing key equipment like the integrated pressure vessel, pressurizer, steam generator, and containment. This reactor type is planned for 100 MWe electrical power / 305 MWt thermal power, with a design life of 60 years and a refueling cycle of up to 24 months, offering the core advantages of long endurance and low operational costs.
In its medium to long-term plan, the company will launch two differentiated products to expand its scenario coverage. The Junhe-2 (50 MWe high-temperature gas-cooled small reactor) is aimed at high-temperature process heat supply and small to medium-scale power generation, suitable for industrial steam, chemical heating, and district heating. The Junhe-3 (5 MWe transportable micro-reactor) is designed for miniaturization, mobility, and rapid deployment, suitable for remote mines, islands, emergency power, and special scenarios. This three-tier product matrix provides comprehensive coverage from "large-scale commercial, to medium-scale heating, to micro-scale emergency," allowing nuclear energy to move beyond the single form of traditional large power plants towards diversified scenarios, flexible deployment, and widespread application.
Product Type | Power | Positioning
Junhe-1 (Small PWR) | 100 MWe / 305 MWt | Data centers, chemical parks, zero-carbon industrial parks
Junhe-2 (HTGR Small Reactor) | 50 MWe | High heat demand scenarios in industrial parks
Junhe-3 (Transportable Micro-reactor) | 5 MWe | Remote areas, special energy scenarios
At an advanced nuclear energy development seminar, Tian Jiashu stated, "The challenge facing global small reactors today is no longer just 'can the technology be built,' but accelerating the transition from technological exploration to more realistic industrialization, driven by policy support, market demand, and regulatory optimization."
What Did Sequoia See?
The "power black hole" of AI data centers is the direct catalyst for this wave of SMR enthusiasm. According to the IEA's "Energy and AI" report, global data center electricity consumption is expected to rise from about 415 TWh in 2024 to about 945 TWh by 2030. The delivery cycle of grid expansion and power generation equipment is much longer than data center construction. What is truly scarce is not just a kilowatt-hour, but a safe, reliable, low-carbon, and sustainably supplied kilowatt-hour. The global SMR market is projected to grow from $6.7 billion in 2025 to $9.9 billion in 2026, a compound annual growth rate of 47.6%, and is expected to reach $92.2 billion by 2032. The IEA predicts that investment in small reactors will need to grow from under $50 billion currently to $250 billion by 2030, potentially reaching $900 billion by 2050. Goldman Sachs forecasts nearly 2 GW of global SMR capacity by 2030, with cumulative deployment reaching about 46 GW by 2045.
The US Department of Energy has invested over $10 billion in new nuclear energy since 2020, including over $4 billion in the ARDP program and $2.7 billion in HALEU enrichment and supply chains. As of February 2025, there were 127 SMR designs globally (29 more than the previous report), with 51 in pre-licensing or formal licensing review. NuScale Power's SMR design has been approved by the US Nuclear Regulatory Commission; Rolls-Royce SMR has strong government support in the UK; and France's EDF is advancing its Nuward small reactor design. Tech giants like Google, Microsoft, and Amazon are all investing in nuclear SMRs to secure energy for their future computing expansion. NVIDIA CEO Jensen Huang invested in TerraPower, and OpenAI's Sam Altman invested in the SMR company Oklo – the urgency for nuclear power from Silicon Valley is now being felt in China.
Recently, China's National Energy Administration convened major nuclear power state-owned enterprises (SOEs) for a special seminar on small reactor development. The State Administration of Science, Technology and Industry for National Defense also held a related seminar. Domestic small reactor R&D has long been led by the three major nuclear power SOEs – CNNC, CGN, and SPIC – along with the Chinese Academy of Sciences, resulting in over ten technical routes of different uses, power levels, and reactor types. The emergence of commercialization companies like Junhe Atomic since 2025 is a significant marker of the acceleration of the small reactor industrialization process.
Globally, the SMR track has formed several mature technical routes, each with varying maturity, deployment pace, and suitable scenarios, with several benchmark companies emerging in China. First, the light water-cooled reactor (including PWR and BWR) route is the mainstream path with the highest commercial maturity and lowest deployment risk. Junhe Atomic is firmly committed to the advanced small PWR route, leveraging a century of PWR commercial experience. The technology system, safety regulations, and supply chain are fully mature, adapted for distributed energy scenarios through miniaturization, modularization, and pump-free design. The core advantages are fast engineering deployment, controllable safety, and significant cost benefits, while the weakness is that the technological iteration tends towards optimization rather than disruption. Domestic representative companies include Junhe Atomic and CNNC's Linglong One team, making it the most suitable solution for AI computing power supply and park-level integrated energy.
Second, the High-Temperature Gas-Cooled Reactor (HTGR) route is a Gen IV advanced nuclear technology, offering high inherent safety and excellent high-temperature performance, suitable for power generation and heat supply, fitting industrial high-temperature process scenarios. The advantages are high inherent safety and scenario diversification, while the weaknesses are high construction costs and slower large-scale deployment. The benchmark domestic project is the Shidaowan nuclear power plant, the world's first commercial HTGR demonstration project, now grid-connected and technically validated, gradually advancing towards smaller modular iterations.
Third, the Liquid Metal Reactor (sodium-cooled SFR, lead-cooled LFR) route is a Gen IV frontier technology, offering high power density, long refueling cycles, and suitability for miniaturization, with enormous long-term potential. However, it involves high technical difficulty, immature supply chains, and insufficient engineering validation, leading to a longer commercialization timeline. Most domestic startups in this track focus on cutting-edge R&D, currently in laboratory and prototype reactor validation stages.
Fourth, the Molten Salt Reactor (MSR) route uses molten salt as fuel and/or coolant, operating at high temperatures. Kairos Power's 140 MWe FHR has received a construction permit from the NRC, making it the first US company to receive a permit for a Gen IV SMR. TerraPower's 345 MWe Natrium project is expected to be operational by 2032. Fifth, the Heat Pipe Microreactor route uses heat pipes for high thermal conductivity. Westinghouse's eVinci (5 MW), Oklo's Aurora Powerhouse, and Radiant Industries' Kaleidos (1.2 MW) are under development.
A JPMorgan report, citing the IEA, suggests that if deployment proceeds smoothly, SMRs could account for 10% of global nuclear power capacity by 2040, with the US contributing 20% of that growth. China's domestic "computing-electricity synergy" policy opens up a trillion-yuan market for computing power infrastructure. SMRs offer power reliability of 99.999% and over 8,000 annual utilization hours, with dedicated lines avoiding grid losses, making them currently the optimal solution for zero-carbon power supply in computing parks.
Of course, Junhe Atomic still has a significant journey ahead before it achieves truly commercial power supply. The company is just over a year old, and the "Junhe-1" has only completed its conceptual design, currently in the optimization phase for core physics, thermal hydraulics, and the main system. Steps like demonstration reactor construction, nuclear safety review, and fuel loading and commissioning have not yet been initiated. The nuclear power industry often takes a decade or more to go from design to commercial operation. This tests not only technical capability but also the patience and long-termism of capital. The hundreds of millions of yuan in Angel funding led by Sequoia Capital China is not buying near-term cash flow; it is investing in Tian Jiashu's team's "all-in-one" capability for design, construction, review, operation, and project development, as well as the certain demand anchor for the small PWR route in the AI computing era. The investment logic for the nuclear fission SMR track is clear: nuclear fusion represents the future, while SMRs are addressing the practical needs of the next decade. In this race, the winner will not be the one with the flashiest technology, but the one that first transforms a "non-standard civil engineering project" into a "standardized industrial product."