The Sun
in a Bottle

Section I

The Moment

At 1:03 AM on December 5, 2022, at the National Ignition Facility inside Lawrence Livermore National Laboratory in California, 192 laser beams fired simultaneously into a gold cylinder the size of a pencil eraser. Inside that cylinder — called a hohlraum — the beams converted into a bath of soft X-rays. The X-rays struck a plastic capsule roughly two millimetres in diameter, containing a frozen shell of deuterium and tritium — two isotopes of hydrogen. The capsule imploded. The fuel compressed. And for a fraction of a second, in a space no larger than a peppercorn, the temperature and pressure reached conditions found only at the cores of stars.

The result: 3.15 megajoules of fusion energy released from 2.05 megajoules of laser energy delivered to the target. More energy out than in. Fusion ignition — the moment a self-sustaining fusion reaction produces more than it consumes — had been predicted theoretically for decades. It had been pursued by the world's largest laser system for more than 60 years of collective research effort. On that December morning, it happened for the first time in human history.

An important caveat entered the conversation immediately, and it is worth holding clearly in mind: the energy produced by the fusion reaction is larger than the laser energy delivered to the hohlraum — but it is far smaller than the total energy consumed by the NIF laser system. The facility's 192 lasers require roughly 300 megajoules of wall-plug electrical energy to fire. The 3.15 MJ of fusion energy represents about 1% of what the facility consumed. "Ignition," as defined here, refers to the energy balance at the target — not a commercially viable energy exchange. The physics has been demonstrated. The engineering challenge is what comes next.

Section II

The Science

Nuclear fusion is the process that powers stars. In the Sun's core — where temperature reaches 15 million degrees Celsius and pressure exceeds 340 billion atmospheres — hydrogen nuclei fuse into helium, releasing energy according to Einstein's E=mc²: a tiny amount of mass converts into an enormous amount of energy. The Sun converts roughly 600 million tonnes of hydrogen to helium every second. Humanity's ambition is to replicate this process at a useful scale, under controlled conditions, on a planet that provides neither stellar gravity nor stellar pressure.

The approach used by NIF is called inertial confinement fusion (ICF). Rather than trying to contain plasma in a magnetic field (the approach of ITER and most other fusion reactors), ICF uses extreme laser power to compress fuel so rapidly that the fusion reaction occurs before the material can fly apart. The fuel is held together by its own inertia — hence the name. The sequence takes less than a nanosecond.

The deuterium-tritium fuel combination was chosen because it has the lowest ignition threshold of any fusion reaction available to experiment — though tritium is rare and must be bred from lithium, which adds supply chain complexity. Future power plants may eventually use the more abundant deuterium-deuterium reaction, which has a higher ignition threshold but uses seawater-derivable fuel with essentially limitless reserves.

The product of D-T fusion is a helium nucleus (an alpha particle) and a high-energy neutron. The neutron carries most of the energy — roughly 80% — and does not interact easily with electromagnetic fields, which is why containing and capturing it for power generation requires the surrounding reactor structure to absorb it as heat. The alpha particle, by contrast, can be used to heat the remaining fuel and sustain the reaction. A burning plasma — one where alpha heating dominates over external energy input — was the precise threshold crossed at NIF on December 5, 2022.

Section III

The Record

Ignition is not a binary switch. The December 2022 shot was a threshold crossing, not an endpoint. The NIF team continued refining capsule design, laser pulse shape, and target configuration, conducting additional ignition shots in the months and years that followed.

Four further ignition shots were confirmed through early 2024: July 30, 2023; October 8, 2023; October 30, 2023; and February 12, 2024. Each successive experiment produced the same fundamental result — more fusion energy out than laser energy in — while the team worked to improve yield. The February 2024 experiment produced an estimated 5.2 megajoules of fusion energy from 2.2 megajoules of laser input: roughly double the energy from the first ignition shot and a target gain of approximately 2.4.

The improvement trajectory matters enormously. From 3.15 MJ in December 2022 to 8.6 MJ in April 2025 — a gain of nearly 2.7× in fusion yield — demonstrates that ICF ignition is not a one-time anomaly. It is a reproducible phenomenon, subject to the ordinary laws of engineering improvement. Better capsule fabrication, optimised pulse shaping, and refined hohlraum geometry have each contributed to the yield increase. The physics is stable. The engineering curve is steep in the right direction.

The wall-plug efficiency gap — the 300 MJ consumed to produce 8.6 MJ of fusion — remains the central engineering challenge for commercial power. That ratio needs to improve by roughly a factor of 50 to reach breakeven at the facility level, and then continue improving to reach commercial viability. No one in the scientific community expects NIF itself to become a power plant. What it has demonstrated is that the physics works. That proof changes the calculus for everyone building what comes next.

Section IV

The Race

NIF is a science facility, not a prototype power plant. The path from ignition physics to commercial electricity requires a generation of engineering innovation that neither LLNL nor any government laboratory is designed or funded to deliver alone. Three parallel tracks now define the race to commercial fusion: the international government programme, and the two most credible private-sector contenders.

ITER — the International Thermonuclear Experimental Reactor, under construction in southern France — is a magnetic confinement fusion experiment funded by 35 countries including the United States, European Union, China, India, Japan, Russia, and South Korea. ITER uses a different approach from NIF: superconducting magnets confine a plasma inside a doughnut-shaped chamber called a tokamak. ITER aims for first plasma around 2035 and sustained fusion demonstration by the late 2030s. It is not designed to generate electricity — that is the job of DEMO, a follow-on reactor not yet in construction. Commercial magnetic confinement fusion power, on this timeline, is a 2040s or 2050s proposition at best.

Helion Energy, backed by Sam Altman and contracted with Microsoft to sell electricity from its first commercial plant by 2028, uses a pulsed field-reversed configuration — a hybrid approach between ICF and magnetic confinement. In February 2026, Helion's Polaris prototype reached plasma temperatures of 150 million degrees Celsius and became the first privately developed fusion machine to operate with actual deuterium-tritium fuel. Its commercial Orion plant, under construction near Malaga, Washington, targets 50 megawatts of generation. Microsoft has a 500 MW follow-on agreement. The target date is 2028.

Private fusion investment globally exceeded $7 billion by 2025, concentrated in roughly 40 startups across the United States, United Kingdom, Germany, Canada, and Australia. The field has moved from being the domain of government physics laboratories to a competitive commercial landscape — a shift driven precisely by the NIF ignition result, which removed the last scientific objection to the enterprise. The argument that fusion was always "30 years away" rested on the claim that ignition itself was unproven. That claim is no longer available.

The civilisational stakes are clear enough. Fusion fuel — derived from lithium and seawater — is effectively inexhaustible on any human timescale. A successful fusion economy would produce no greenhouse gases during operation, generate orders of magnitude less radioactive waste than fission, carry no risk of chain-reaction meltdown, and be available to every nation regardless of geography. Whether the engineering gap gets closed in the 2030s or the 2050s is the remaining question. December 5, 2022 established that it will eventually be closed at all.