A cylinder of spinning liquid lithium, collapsed by pistons once a second – and a single scientific milestone the whole plan hangs on.
Josh Nycholat, Manager of Investor Relations at General Fusion, opened with the company’s defining premise. Fusion energy has not been commercialized because the industry treated it primarily as a physics problem. His company, by contrast, approaches it as an engineering one.
General Fusion is Canada’s leading fusion technology developer. Since 2002, it has pursued a route called magnetized target fusion, building, operating and testing machines that advanced the technology step by step.
Energy Tech Summit 2026 keynote speaker Josh Nycholat, Manager of Investor Relations at General Fusion.
The company at a glance
The founding insight came from Canadian physicist Dr. Michel Laberge, who recognized a significant gap in why fusion had not reached commercialization. Consequently, he argued for addressing those challenges from the ground up, using existing technologies and existing materials wherever possible to put practical electrons on the grid cost-effectively.
How fusion works, briefly
For anyone unfamiliar, Nycholat started from first principles. Fusion is the process powering the sun and the stars – the nuclear process the universe runs on. It brings two positively charged atoms together under tremendous pressure and density, releasing energy as they combine.
Because Earth lacks the gravity that makes this happen in a star, the same force has to be created inside a machine. The two most common fuels are deuterium and tritium, both hydrogen isotopes. Deuterium occurs naturally, whereas tritium must be produced.
General Fusion therefore takes deuterium and tritium, creates fusion under the right conditions, and captures the resulting energy using a proprietary liquid metal wall – which is where the engineering argument begins.
One wall, four problems
Nycholat named four barriers that have blocked commercial fusion: material degradation, fuel sourcing, energy capture and cost. Notably, the liquid metal wall addresses all four inherently rather than separately.
On material degradation, the wall protects the machine from the reaction itself. When deuterium-tritium fusion occurs, neutrons are released, so the first surface they strike has to withstand sustained bombardment. A liquid wall cannot be degraded the way a solid one would be.
On fuel sourcing, the same lithium does a second job. As neutrons travel through it, they interact with the lithium and produce tritium fuel. Crucially, that happens at a ratio greater than 1.5, which the company’s partners at the UK Atomic Energy Authority have assessed as sufficient to repower the reaction and start up new machines besides.
On energy capture, the wall acts as a heat sink. It creates a hot liquid that can then be pumped through a heat exchanger in the balance of plant.
How the commercial machine will work
The commercial design centres on a large cylindrical vessel holding an inventory of liquid metal. First, the vessel spins fast enough to push that metal outward against the wall, so it fully surrounds the chamber.
Then the plasma – the deuterium-tritium fuel cloud – is injected into the centre. Finally, the liquid metal wall collapses around it.
That collapse happens once per second. Gas-driven pistons push against the metal, starting at the top and bottom so the collapse forms a full sphere. The company has already tested the fluid dynamics of that sequence experimentally, and Nycholat said they are confident the design achieves what it needs to.
Once fusion occurs at the centre, neutrons are released into the liquid metal. As a result, fuel is produced, the metal heats up, and that heat translates into electricity for the grid.
Two decades of de-risking, three core proofs
General Fusion has spent decades de-risking its technology system by system, running a couple of dozen test beds. Those efforts culminated in three main achievements.
The first is plasma performance. Because the approach is pulsed, the plasma has to last long enough to be mechanically compressed – an absolute requirement for the technology to work at all.
The second is liquid compression performance. The wall must collapse with the right symmetry, so that it interacts with the plasma without destabilizing it. A test bed called the cylindrical water compressor validated the cylinder-to-sphere collapse required.
The third is plasma compression, which is where the approach differs from others. The company needed to confirm that mechanically compressing a plasma produces rising neutron yield. Therefore it built a test bed in Vancouver that collapsed a small-scale plasma using force, and the observed yield matched expectations. Those results were published, which gave the company the validation it needed to proceed.
LM26, and the milestone that matters
LM26 was designed and assembled in December 2024 on the basis of those results. Its purpose is narrow but decisive: to demonstrate that a plasma can be mechanically compressed sufficiently to reach 100% Lawson.
Progress has followed quickly. The machine achieved first plasma in February 2025 and first plasma compression in April 2025. Since then, the team has run repeated compressions, improving performance shot by shot while collecting and analysing the data.
Three milestones remain, running through the end of 2028: 1 keV, 10 keV, and 100% Lawson. Each represents fusion yield increasing further. That final one carries the weight, because 100% Lawson is the scientific validation that confirms the approach works and unlocks the next phase.
From validation to a first plant
After LM26 comes a stage the company calls demonstration of commercial systems – essentially the enabling technologies needed to move from scientific validation to a first-of-a-kind power plant.
Those requirements follow directly from the design. Because the machine uses a rotating liquid metal wall compressed by pistons, it needs seals and valves that work with hot lithium. It also needs a heat exchanger capable of interacting with hot lithium, plus a tritium extraction fuel cycle. Some organizations are already working on these technologies, so partnership matters at this stage.
A first-of-a-kind plant follows, delivering the first large-scale liquid compression producing either industrial heat or electricity.
Going public, and why the funding matters
Because of where the technology and the science now stand, the company has decided to enter the public markets through a merger with a special purpose acquisition company – the same group that took a prominent small modular reactor developer public.
The reasoning Nycholat gave was about participation: the public markets should be able to take part in these milestones as they happen.
General Fusion’s representative Josh Nycholat sharing insights on ETS2026 Tech Talks stage.
Takeaway
Most fusion energy pitches lead with plasma physics. This one led with lithium, seals and heat exchangers, because General Fusion’s bet is that the hard part was never the reaction. A spinning wall of liquid metal absorbs the neutron damage, breeds its own tritium, and carries away the heat – three barriers solved by one engineering choice, with cost following from the decision to use materials that already exist. Whether the bet holds gets answered by a single number in 2028. Until 100% Lawson arrives, everything downstream of it, including the 2035 plant, remains contingent.
The keynote stage returns at Energy Tech Summit 2027 in Bilbao, April 7–8.
