Bright spark? The race for fusion energy is no longer science fiction

In the 2008 Iron Man movie, a fictional miniature fusion energy device powers Tony Stark’s metal suit and protects his heart. A much larger version of the Arc Reactor powers the headquarters of Stark Industries.

Fiction could soon become reality.

The energy answer the world is increasingly betting on isn’t more oil, more gas or even more solar panels. It’s fusion, which could prove a vital source in the decades ahead.

For a sector that was always 30 years away from success, suddenly it’s looking like it could be 10. Across the globe, dozens of start-ups and government-funded projects are working on ways to replicate, and harness, the power of the sun to create a clean and almost limitless source of energy and meet growing demand for electricity.

There is rising consumption from industry, electric vehicles and air conditioning. But the one driver that is outpacing all others and has ignited intense competition among the superpowers is data centres.

Global electricity generation to supply them could nearly triple by 2035 from 460 terawatt-hours in 2024, the International Energy Agency (IEA) says. The AI race is, at its core, an energy race. Every AI query has an energy cost. Every time you stream a video, use ChatGPT or a digital service, energy is being consumed in vast, relentless amounts.

The IEA says we’re in the Age of Electricity and forecasts annual global growth of 3.6 per cent between 2026 and 2030.

At this juncture, all that demand is colliding with a turbulent moment in geopolitics. The Iran war has made it loud and clear that dependence on fossil fuels is not just a climate problem but also a security vulnerability.

The impasse on the Strait of Hormuz has tightened supply lines that have led to oil spikes felt across the world. Businesses are feeling the impact and passing the cost to households who are hit by increased day-to-day spending on essentials that include groceries, transport and utlities.

Electricity drives modern life and the scramble to secure it that has never felt more urgent.

Ramping up use of planet-warming fossil fuels isn’t an option because of the growing climate change impacts.

Closer to reality

“There’s a clear demand signal right now that is strong, growing and voracious. There is a need for power,” said Mr Rick Needham, Chief Commercial Officer for US firm Commonwealth Fusion Systems (CFS), which has attracted nearly US$3 billion (S$3.8 billion) in funding, the largest of any US fusion start-up.

That demand is underpinning the race to be the first to build a commercial fusion power plant. Billions of dollars of government and private funds – including Big Tech money – are propelling fusion forward.

“We have a much larger group of companies developing a diverse range of more advanced technologies, and together that has also built out the wider supply chain, crowding in investment and talent,” said Dr Samuel Ward, senior consultant at Helixos, an advisory firm focusing on advanced nuclear and fusion power.

Big technical challenges still lie ahead to achieve commercial and affordable fusion power.

But success could transform the global energy sector and the geopolitics of energy, shaking up the stranglehold fossil fuel producers have on the global economy.

“Fusion is not just a new energy technology. Fusion is a fundamental change in how we interact with energy,” said Mr Andrew Holland, chief executive of the Fusion Industry Association, a Washington-based global advocacy group.

The current energy system is dominated by geography – materials that are extracted from the ground burned or resources such as minerals. Fusion reactors could be built just about anywhere and be fuelled by some of the most abundant resources in the universe.

100 million deg C

Fusion energy combines light-weight atomic nuclei (typically hydrogen isotopes, deuterium and tritium) to form heavier ones, releasing massive amounts of energy without long-term radiation or risk of nuclear meltdowns. It is different from fission energy used in big nuclear power plants, which involves splitting atoms and releasing large amounts of radiation.

But harnessing fusion energy involves machines that can withstand temperatures of 100 million deg C or more and immense pressures for long periods.

The idea is to create self-sustaining fusion reactions within a super-hot plasma and capture the heat to generate electricity.

It sounds like science fiction. Yet, fusion reactors exist and have created plasma for increasing lengths of time – 1,337 seconds is the current record by French and Chinese reactors. But creating and holding the plasma takes energy. The aim is to create reactors that produce more energy than they consume, something that’s proven to be possible but hard to achieve.

CFS is confident its SPARC demonstration plant, which is already 75 per cent complete, will produce more energy than the plant consumes by 2027.

The firm, backed by Singapore’s Temasek, Nvidia and others, plans to quickly take the lessons learnt from the US$1 billion SPARC and start building its ARC commercial power plant in Virginia, with completion in the early 2030s. Google, also an investor, has already signed a deal to take half the plant’s electricity.

CFS has plenty of competition.

China eyes the prize

In Hefei, China’s BEST (Burning Plasma Experimental Superconducting Tokamak) is taking shape quickly, with completion also set for 2027. It is backed by the Chinese Academy of Sciences’ Institute of Plasma Physics and start-up NeoFusion. The aim is to achieve net fusion power gain and to demonstrate fusion-based electricity generation by 2030, potentially beating ARC’s timeline to produce power.

Canada’s General Fusion, also backed by Temasek, is perfecting its own plant design with the aim for a commercial power plant around 2035.

Globally, there are 53 private-sector fusion start-ups, Mr Holland said, with about 30 in the United States, a small but growing number in China, plus others in Europe, Japan, Australia and New Zealand. Just over US$10 billion in private funding has gone to the sector, most of it since 2021.

Government funding is even larger through industrial support programmes and multinational research initiatives. China has dramatically scaled up its fusion industry and investment in the past five years. The government’s 15th Five-Year Plan (2026-2030) promises “extraordinary measures” to secure breakthroughs in fusion energy and positions the sector as a key future industry.

Despite the competition and fractured global geopolitics, the key players are also part of the world’s largest fusion experiment: the International Thermonuclear Experimental Reactor, or ITER. China, the European Union, India, Japan, South Korea, the Russian Federation and the United States are the key funders of the giant magnetic confinement fusion device being built in France.

The goal is to demonstrate the feasibility of fusion power at an industrial scale by 2033-2034 – with the current price tag ballooning to more than €20 billion (S$29.7 billion).

Challenges ahead

Despite the rapid advances, plenty of engineering challenges remain.

These include coping with the extreme heat, materials strong enough to withstand the intense bombardment from neutron particles produced by the fusion reactions and creating and harvesting tritium fuel to feed and sustain the plasma.

The most developed and oldest design is the tokamak, which uses powerful magnetic fields to confine the plasma in a donut-shaped chamber called a torus. The magnets help control the position and shape of the plasma to keep it away from the walls. Powerful computer models and AI help simulate how the plasma will behave.

The walls need to be able to withstand not only the extreme heat of several thousand deg C but also the neutron bombardment, which is also a risk to the magnets that surround the torus.

Tungsten, copper and stainless steel are among the materials being trialled by developers to withstand the heat. But neutrons are likely to degrade the lining.

One idea is to design the parts of the wall to be easily replaced during regular maintenance.

“It’s something we’ve been thinking long and hard about. We’ve actually kind of incorporated it into the design of the plant itself,” said Mr Needham.

There’s also the challenge of creating a constant flow of tritium fuel.

Tritium is a scarce, radioactive isotope of hydrogen that is produced in small quantities by fission nuclear reactors. Various fusion reactor designs include ways to generate tritium. By contrast, deuterium is much more abundant and can be harvested from sea water.

SPARC will trial the use of a molten salt blanket, a layer directly behind the reactor, or vacuum vessel, wall. A key component of the salt is lithium, which produces tritium when hit by neutrons. The tritium will be harvested and fed back into the vacuum chamber in a self-perpetuating cycle. The salt blanket will also protect the magnets from neutron radiation and absorb the heat from the plasma chamber.

General Fusion has created a liquid metal wall, which is designed to absorb neutrons, protect the machine from fusion damage, breed tritium fuel and capture heat.

Much is riding on the success of these innovations. For CFS, questions remain whether it can meet its very aggressive timeline to start building the 400MW ARC as soon as 2027 while also incorporating engineering updates as construction proceeds.

“We expect ARC to be online in the early 2030s. There are things moving in parallel, but this is what you do if you want to bring fusion power plants online as soon as you can,” Mr Needham said.

More broadly, questions on affordability of the power generated, competition with renewables and public acceptance also remain potential challenges for fusion energy.

Renewable energy, such as solar and wind, plus battery storage is being rolled out at breathtaking scale globally and is now cheaper than fossil fuel-generated electricity. In 2025, renewable power made up almost half of the world’s electricity capacity after a record ‌increase in solar installations, according to the International Renewable Energy Agency.

Singapore is watching developments closely but is keen to be a global player.

At Nanyang Technological University, Professor Xavier Garbet is training young scientists in plasma energy to prepare them for the age of fusion power. A key part of his work involves using powerful computers to model plasma behaviour and is working on how AI can greatly speed this up.

A*STAR is focusing on material sciences for fusion to position Singapore as part of the global supply chain. It is also collaborating on fusion plasma modelling at NTU.

“It’s going to be tough. 2040 might be feasible,” said Prof Garbet, who is the Temasek Chair in Clean Energy, when asked if commercial fusion power is feasible within a decade.

Will geopolitics get in the way?

Not surprisingly, Mr Holland and Mr Needham think the US is winning the race but are wary of China and other nations.

“We think the US is still ahead. We think we’re pretty far in the lead,” Mr Needham said.

The US needs more government support and funding beyond what is already being committed, currently about US$800 million a year.

“The government needs to put its thumb on the scale, because if it doesn’t, this is going to be invented in the United States and commercialised somewhere else,” said Mr Holland, whose association in 2025 called for a one-time US$10 billion injection of US public funding into the private fusion industry.

“Without new funding and a decisive shift toward technology development and commercial demonstration, the US will lose the fusion race to competitors like China,” the association said in a position paper.

China clearly wants to win the race and is approaching fusion energy the same way it approached the clean-tech sector: domination through targeted national and provincial support for private companies to drive innovation, quickly build scale and create intense competition between companies.

The US is more private sector-led, driven mostly by funding from tech giants, philanthropy and investment funds.

For the moment, US-China rivalry in AI, semiconductors and trade fights over rare earths and critical minerals aren’t a roadblock for fusion energy, observers say. But that could change as the sector ramps up.

While competition is healthy, there is a risk the fusion race could descend into another front in the US-China trade war. This risks losing sight of the formula for success to date, which has been built on a much broader church of collaboration.

“Fusion needs the urgency that comes with commercial competition but a continued sharing of standards, public institutions and fundamental science research. It’s this type of hybrid arrangement in place today that launched a competitive industry,” said Dr Ward.

In a world hungry for energy and rising risks from the climate impacts of burning fossil fuels, there’s an urgent need for a new source of power.

Fusion could well be the spark that grows to power humanity’s next phase of growth and development. We’ll soon find out if fusion’s great promise is real.