At a recent fusion energy forum, a provocative question was posed to three startup founders: if they were investors, which competing technology route would they least dare to fund, and why? The panel featured Tan Yi of Startorus Fusion, Xu Min of East Rise Fusion, and Wei Xishuo of Honghu Future Energy, each championing a distinctly different approach to achieving practical fusion power. Tan Yi's initial response, asking for permission to rebut before answering, drew laughter as he quipped that the debate might become "completely irreconcilable."

The three companies represent three divergent fusion technology paths: spherical tokamaks, deuterium-helium-3 fuel, and stellarators. Behind their presentations lies a long-running, high-stakes contest among different technical approaches, each with passionate advocates and significant vulnerabilities. Fusion energy carries humanity's vision for an ultimate power source, yet the road to that destination remains frustratingly unclear, with no single route yet emerging as the definitive winner.

According to the latest data from the International Atomic Energy Agency (IAEA), there are currently 174 fusion devices worldwide in operation, under construction, or planned. Of these, 78 are tokamaks, 28 are stellarators, 20 are laser inertial confinement devices, and the remaining 48 follow other technical approaches. The tokamak remains the undisputed mainstream with the deepest engineering experience and closest proximity to commercialization, though its cost and complexity remain daunting. Stellarators, a magnetic confinement cousin to tokamaks, long received far less attention and funding, languishing as a niche approach until recent breakthroughs in superconducting materials and 3D printing dramatically accelerated their development and brought them back into view for commercial fusion ventures.

Deuterium-helium-3, meanwhile, represents both the most ideal fuel and the most distant dream. It requires temperatures far exceeding those needed for deuterium-tritium fusion, and natural helium-3 is virtually nonexistent on Earth. While the Moon's surface holds abundant helium-3 reserves, its extraction and transport remain science-fiction-level challenges. On a track rife with variables, betting on a single technology would be a gamble of enormous proportions, while spreading investments across all options risks diluting resources and losing critical momentum for breakthroughs. Shanghai's choice, however, has been to hold every possibility firmly in hand: regardless of which route ultimately reaches the finish line, the city ensures it stands at the forefront of the race.

While opinions diverge on the best path, there is unanimous agreement on one point: for fusion energy, which appears endless and perfect in theory, no single technical route is flawless. "Every choice right now is a difficult one," Tan Yi admitted. Startorus Fusion has charted a distinctly differentiated path: building on high-temperature superconducting strong-field spherical tokamaks with multi-pulse repetitive operation, using magnetic reconnection to heat plasma. This abandons the traditional approach of maintaining high-temperature plasma for extended periods, opting instead for an internal-combustion-engine-like model of rapid, short-pulse ignition with repetitive work cycles — all with a singular goal of making fusion reactors smaller, simpler, and dramatically cheaper to build.

Short pulses reduce the risk of major plasma disruptions inherent in conventional long-pulse tokamaks, but they introduce new challenges of their own. "The trade-off is that the tokamak operates cyclically, subjecting all magnets and structures to periodic stress, while the first wall faces periodic neutron and thermal loads. How do we ensure materials maintain long service lives under repeated stress and thermal cycling?" Tan Yi drew an analogy to refractory bricks in blast furnaces: they withstand constant two-to-three-thousand-degree temperatures without issue, but after just a few cooling cycles, they crumble into slag. "Until a fusion reactor actually runs, it's very hard to declare any single approach absolutely superior," he said. "Everything ultimately comes down to actual operation and real testing."

Tan Yi acknowledged the costs of short pulses, and Xu Min's assessment is equally candid. "There is no route with only advantages and no drawbacks, just as there is no route with only drawbacks and no advantages — everything is a balance of trade-offs." East Rise Fusion, incubated by Xu Min's team at Fudan University, has chosen the deuterium-helium-3 path. It is the world's first magnetic confinement fusion company using a tokamak configuration with this fuel. Mainstream deuterium-tritium reactions generate copious high-energy neutrons, necessitating massive radiation shielding that inflates device scale and limits siting options. The deuterium-helium-3 route produces almost no neutron radiation, generates minimal radioactive waste, and offers higher energy conversion efficiency — a recipe often described as "perfect fusion."

"By choosing deuterium-helium-3, we avoid the demanding requirements of neutron irradiation-resistant materials. But the trade-off is clear: we must have high-temperature superconducting strong magnets as a prerequisite for the tokamak. Without those, deuterium-helium-3 fusion is economically unviable," Xu Min said. He acknowledged the path is a risky one. Deuterium-tritium fusion will inevitably be the most easily achievable fusion scheme globally; East Rise's positioning is to serve as an important complement to the national mainstream route. "We hope this path can succeed, serving as a Plan B when the nation makes its strategic choices."

That "Plan B" might well be the most treacherous of all. Tan Yi was notably cautious about deuterium-helium-3's prospects: "If deuterium-tritium is this difficult, deuterium-helium-3 demands considerably more. I can hardly imagine designing a fusion reactor that meets power generation requirements using deuterium-helium-3, even with sufficient funding." His hesitation should not be mistaken for conservatism. Every step forward in fusion research tests the physical limits of existing engineering materials.

Wei Xishuo's Honghu Future Energy is betting on stellarators. "I judge whether a route can become a future commercial fusion reactor based on two criteria: first, whether its advantages can be proven experimentally; second, whether its challenges can be resolved within the foreseeable future." He cited experiments at Germany's W7-X, the world's largest stellarator fusion device: "In 2024, W7-X demonstrated that optimizing magnetic field geometry can reduce neoclassical transport below turbulent transport levels, placing stellarators on equal footing with tokamaks. In 2025, the device achieved continuous stable plasma operation for 43 seconds, setting a new fusion triple-product record at that timescale." The constraint that ended that run was primarily engineering-related; after upcoming upgrades, he noted, "high-parameter steady-state operation times will extend further. These experimental results have boosted my confidence in stellarators."

As for the traditional challenges facing stellarators — three-dimensional magnetic field design and complex shaped coil manufacturing — Wei believes technology is dissolving these obstacles: "AI tools have driven significant progress in magnetic field design, and advances in three-dimensional high-temperature superconducting coils, along with declining costs, are very real developments." Critically, substantial national investments in first-wall and blanket material research, as well as neutral beam injection heating and diagnostic equipment developed for tokamaks, can be directly transferred to stellarators, avoiding redundant development. Yet critics point to a fundamental drawback: the three-dimensional twisted coil structure makes stellarators "thick-skinned with a small core" — for the same device scale, their power-generating volume is far smaller than a tokamak's, which would directly inflate future cost per kilowatt-hour.

Different roads, same perilous peaks. The spherical tokamak pursues miniaturization but must withstand the tearing forces of cyclic stress; deuterium-helium-3 is tantalizing but hinges on the risky leap of high-temperature superconducting strong magnets; the stellarator has risen from behind but ultimately contends with its own extreme structural complexity.

"Choosing a technology path isn't just about looking from the present toward the future; it's about looking from the future back to the present," Xu Min observed. He is highly optimistic about the cost-reduction curve of the high-temperature superconducting industry: "I've heard tape manufacturers have pushed material costs down to RMB 40 per meter, down from RMB 400-500 per meter — this is the result of industrial development. Shanghai's high-temperature superconducting industry can fully mature within the next decade, providing very stable, reliable, and economical magnets."

"Fusion technology has not yet reached its final closed state; all routes deserve attempts and technical reserves," Wei Xishuo said. Mi Lei, the investor, offered his answer: invest in all of them. "Fusion parameters are too numerous — every domain and every technical parameter may touch the current ceiling of human technology, whether in materials or devices. Today's ceiling is not tomorrow's ceiling." He continued: "Many technological breakthroughs depend on whether investors are willing to support entrepreneurs in pushing past those limits." Mi added that fusion investment risk is shrinking dramatically as global technology trends become increasingly clear. With capital and talent accelerating into the field, the probability of success rises and commercialization timelines compress. In his view, exploration across different technical routes all serve as beneficial supplements to the national mainstream path, but the stakes extend well beyond that: each route represents China's race against global competitors, ultimately determining whether the nation can achieve breakthroughs and leadership in the next wave of technological revolution.

Capital flowing into fusion has never been more turbulent. Guo Yumin, division director at the Ministry of Science and Technology, reported at the same forum that, guided by central-local policy coordination, domestic fusion venture investment has been exceptionally active this year. In the first half of the year alone, preliminary statistics show 30 primary market financing events — an eightfold year-on-year increase — with total financing reaching RMB 8 billion, a more than 233-fold increase. Nearly 30 startup companies received capital support during the period, while upstream and downstream industry chain enterprises collectively raised approximately RMB 1.5 billion. "Fusion investment now exhibits dual characteristics of large-ticket and consecutive funding, with single-round financing reaching record highs. Long-term confidence in high-quality targets is strengthening, and the investment ecosystem is becoming increasingly diversified," Guo said.

But is fusion investment overheated? Far from it. Guo cited the US Fusion Industry Association's "Global Fusion Industry Report 2026" and European fusion energy organization data: between July 2025 and June 2026, global annual new financing in the fusion industry surged 69% to $4.48 billion, with cumulative industry financing reaching $14.24 billion. From 2022 to 2025, the five major global economies invested a total of €20.35 billion in fusion, with US public funding around €4.95 billion and private funding approximately €2.41 billion. China's public funding reached approximately €5.45 billion — the highest among the five major economies — but private funding was a mere €1.5 billion. This structural gap underscores that China's fusion investment is far from saturated, requiring more market capital to take up the baton.

"Global fusion has entered an era of Sino-American duopoly," Guo said. "How to guide more venture capital to continue entering the fusion field, and how to win the initiative in a fusion race that could determine national destiny, are questions requiring deep thought from the scientific community, industry, and investment circles alike."

Hot capital, however, cannot alter the fact that this is a marathon. The timeline for fusion commercialization is being pushed back across the board. The "Global Fusion Industry Report 2026" shows that among 45 fusion companies providing commercialization schedules, the number planning commercial operations between 2031 and 2040 has declined compared to the 2025 survey, while those targeting 2041 to 2050 have grown significantly. As the global fusion community reassesses commercial expectations, the length of patience will determine who holds the voice in the next phase of capital deployment. And Shanghai appears determined to run to the very end.

Shanghai has already become one of the world's cities with the highest density of fusion devices and the most diverse technology routes. Zhai Jinguo, deputy director of the Shanghai Municipal Science and Technology Commission, offered concrete figures: cumulative fusion industry financing in Shanghai has approached RMB 18 billion, positioning the city to become a global fusion innovation hub with the most complete technology coverage, the strongest financing atmosphere, and the highest talent density. In May this year, Startorus Fusion completed a RMB 500 million Series A+ round, pushing cumulative financing past RMB 2 billion and valuing the company above $1 billion — crossing into unicorn territory. Startorus is far from alone: Shanghai has formed an ecosystem where the national fusion team coexists with multiple unicorns, and the key force underpinning this landscape is the Shanghai state-owned capital system.

Dai Minmin, deputy secretary of the Party Committee and president of Shanghai State-owned Capital Investment Co., Ltd. (Shanghai Guotou), stated that the company has cumulatively invested over RMB 1 billion in the fusion industry, comprehensively covering upstream materials, core components, multiple technology routes, and the "AI for Fusion" segment across the full industry chain, leveraging RMB 10 billion in social capital into the fusion sector. In May 2024, Shanghai Guotou launched its forward-looking layout in the fusion industry. At that time, fusion had not yet been written into the national 15th Five-Year Plan framework, and the broad "fusion fever" had not yet ignited. Over the following year, Shanghai Guotou brought in China Fusion Energy Co., Ltd., identified 30 scientist teams, and systematically constructed a full industry chain landscape for fusion energy.

Beyond introducing the national team at China Fusion and incubating tokamak-route companies like Startorus and East Rise, Shanghai Guotou has conducted in-depth research on multiple non-mainstream paths including laser fusion, stellarators, and field-reversed configurations. Through competitive selection mechanisms, it has reserved capacity for several fusion device projects, including the double-cone collisional ignition laser fusion project led by Shanghai Jiao Tong University's academician Zhang Jie, the high-temperature superconducting stellarator project led by Lin Zhihong's team (Honghu Fusion), and the field-reversed configuration project led by Sun Xuan's team at the University of Science and Technology of China (Xingneng Xuanguang).

Early positioning and multi-path validation have given Shanghai what is now arguably the most complete and systematic controllable fusion track layout in the country. Starting early tests strategic vision, but the fusion marathon tests endurance more than starting speed. In 2024, the Shanghai Future Industry Fund launched under Shanghai Guotou's umbrella (initial scale RMB 10 billion, later expanded to RMB 15 billion), with Shanghai's fiscal system as the single sole limited partner. Liu Lipeng, executive director of investment at the Shanghai Future Industry Fund, told The Paper that the fund established a "15+3" fund cycle from day one. In the past two years, many provinces have extended government fund durations for future industries, but two years ago, an 18-year ultra-long duration patient capital fund was rare nationwide. The Future Industry Fund operates with an "80% sub-funds, 20% direct investment" structure for comprehensive allocation while remaining pluralistic and inclusive on technology route selection.

"Shanghai is already the city with the highest density of fusion devices in the world — not even one of them," Liu Lipeng said. "Once a device cluster forms, industry chain elements like tapes, magnets, and core components naturally aggregate. This creates a naturally formed symbiotic ecosystem: devices attract the industry chain, the industry chain feeds back into devices, and this gives rise to world-class industrial systems and attracts world-class talent." Liu believes that "we must maintain strategic agility while absolutely avoiding strategic missteps." Controllable fusion, as an ultimate energy source with enormous potential, has yet to see its technical routes fully converge; both Chinese and American investment circles remain open to multiple approaches. Before technical convergence, excluding any single route may mean missing the final outcome. Breakthroughs in high-temperature superconductors and the deepening application of AI are injecting new variables into the race among different routes.

"Controllable fusion is a long-distance race with a known finish line, and the global trend is toward multiple parallel routes," said Chen Rui, CEO of Startorus Fusion. "While over 50% of devices are focused on tokamaks, I believe even at the finish, tokamaks may not dominate alone. Shanghai's advanced philosophy of pursuing multiple technical routes simultaneously represents strategic wisdom in the fiercely competitive international fusion race." Multiple fusion companies landed in Shanghai in 2025 alone. "If controllable fusion is truly achieved, the Yangtze River Delta will become the heartland of the fusion industry — currently it is the only high ground," Chen told The Paper previously.