Nuclear fusion: the technology about to change everything

01The fusion breakthrough explained

Nuclear fusion is the reaction that powers the stars — two light atomic nuclei combine to form a heavier nucleus, releasing enormous energy from the conversion of mass. The difference in mass between the reactants and products is manifested as the release of energy, following Einstein's E = mc². Active stellar cores are powered by fusion, and nucleosynthesis via fusion in the Big Bang and in stars creates all elements lighter than nickel.

On Earth, the most promising fusion reaction for energy production is the deuterium-tritium (D-T) reaction, which produces helium and a neutron, releasing 17.6 MeV of energy per reaction. Deuterium is abundant in seawater; tritium can be bred from lithium. The fuel supply is effectively limitless — one gram of fusion fuel yields the energy equivalent of roughly 24 tonnes of coal. The challenge has never been about fuel scarcity; it has been about achieving and sustaining the extreme conditions needed for fusion to occur.

02Magnetic confinement vs inertial confinement

Two major approaches dominate controlled fusion research. Magnetic confinement fusion (MCF) uses magnetic fields to confine fusion fuel in the form of a plasma, keeping the superheated gas away from reactor walls while it reaches temperatures of over 100 million degrees Celsius — ten times hotter than the sun's core. The leading magnetic confinement device is the tokamak, a donut-shaped chamber first developed in the Soviet Union in the 1950s.

Inertial confinement fusion (ICF), by contrast, uses high-energy lasers to compress a small fuel pellet to extreme densities, triggering fusion in a brief implosion. The National Ignition Facility (NIF) at Lawrence Livermore National Laboratory demonstrated this approach in December 2022, achieving the first controlled fusion ignition — a reaction that released more energy than the lasers deposited into the target. Stellarators, a third approach championed by Germany's Wendelstein 7-X, use complex twisted magnetic fields to confine plasma without the instabilities that plague tokamaks.

03The Q-factor milestone and what it means

The key metric for fusion energy is Q — the ratio of fusion power output to external heating power input. A Q of 1 means "scientific breakeven," where the fusion produces as much energy as is put in. A Q above 1 means net energy gain. The NIF's December 2022 result achieved Q ≈ 1.5 in terms of laser energy to fusion energy, a historic first. In 2023, NIF improved this to Q ≈ 2.4, meaning the fusion output was more than double the laser input.

However, the Q-factor can be misleading. The NIF result measures energy from the laser hitting the target, not the total energy drawn from the electrical grid to power the laser, which was hundreds of times larger. Engineering Q — net electricity out versus electricity in — remains the true benchmark, and no fusion device has achieved it yet. ITER, the international tokamak under construction in France, is designed to reach Q = 10, producing 500 MW of fusion power from 50 MW of heating input, but its first plasma is not expected until the late 2030s.

04Private fusion companies and timelines

While government mega-projects like ITER proceed on decadal timescales, a wave of private fusion startups is racing to commercialise fusion much faster. Over 50 private fusion companies have been founded globally, attracted by breakthroughs in high-temperature superconducting magnets, advanced plasma diagnostics, and machine learning for plasma control. Many aim for commercial fusion in the 2030s, a timeline that ITER's schedule would not meet until the 2040s or later.

The leaders include Commonwealth Fusion Systems (CFS), spun out of MIT, which has raised over $2 billion and is building the SPARC tokamak using REBCO superconducting magnets to achieve a compact device. Helion Energy, backed by Sam Altman, uses a field-reversed configuration to compress plasma directly for electricity generation, aiming for a 50 MW demonstrator by 2028. TAE Technologies uses a beam-driven field-reversed configuration with boron-proton fuel — a reaction that produces no neutrons. Tokamak Energy builds compact spherical tokamaks with HTS magnets.

05Materials science challenges for fusion

The plasma physics milestones grab the headlines, but fusion's hardest engineering problems may be in materials science. A fusion reactor's first wall faces a brutal environment: neutron bombardment at 14 MeV, temperatures exceeding 1000 degrees, and intense magnetic and mechanical stresses. No existing material has been demonstrated to survive these conditions for the years of continuous operation that a power plant requires.

Neutron damage is the most severe problem. Each 14 MeV neutron from the D-T reaction displaces dozens of atoms in the wall material, causing embrittlement, swelling, and transmutation. Tungsten and advanced steel alloys like EUROFER97 are candidates, but testing them requires a fusion-relevant neutron source that does not yet exist. The IFMIF-DONES facility, under construction in Spain, aims to provide such a source but will not be operational until the late 2020s. Tritium breeding is another challenge — reactors must produce their own tritium fuel from lithium blankets lining the wall, and no full-scale breeding blanket has been demonstrated under real fusion conditions.

06Fusion vs fission: safety and waste

Fusion's appeal over nuclear fission rests on three pillars: safety, waste, and fuel supply. A fusion reactor cannot melt down — if confinement is lost, the plasma simply cools and the reaction stops. There is no chain reaction, no critical mass, and no risk of runaway. The fuel inventory in the reactor at any given time is measured in grams, not tonnes, so even a catastrophic breach would release a trivial amount of radioactive material.

On waste, fusion is dramatically cleaner than fission. The D-T reaction produces only helium and neutrons. The radioactivity comes from neutron activation of reactor components, which produces low- and medium-level waste that decays to safe levels within 100-200 years — compared to fission's high-level waste that remains dangerous for millennia. Fusion fuel — deuterium from seawater and lithium — is effectively unlimited, eliminating the geopolitical and supply-chain concerns that accompany uranium mining and enrichment.

07When will fusion power the grid?

The honest answer is: nobody knows for certain, but the timeline is shortening. ITER is targeting first plasma in the late 2030s and DT operation in the 2040s. Commonwealth Fusion Systems aims for a demonstration plant producing net electricity by the early 2030s, leveraging compact tokamaks enabled by high-temperature superconducting magnets. Helion Energy has committed to delivering electricity to Microsoft by 2028 — the first commercial fusion power purchase agreement ever signed.

Whether these timelines hold depends on overcoming the materials, tritium breeding, and engineering challenges that have historically caused every fusion timeline to slip. But the combination of private capital, advances in superconducting magnets, AI-driven plasma control, and demonstrated net energy gain means that fusion is closer than it has ever been. The question is no longer whether fusion can produce net energy — it can. The question is whether it can do so reliably, cheaply, and at scale.