General Fusion Inc. is a Canadian company based in Richmond, British Columbia, which is developing a fusion power technology based on magnetized target fusion (MTF). The firm was founded in 2002 by Dr. Michel Laberge. , it has more than 150 employees.
The technology under development injects a magnetized target, a plasma mass in the form of a type of plasmoid termed a compact toroid, into a cylinder of spinning liquid metal. The target is mechanically compressed to fusion-relevant densities and pressures, by anywhere from a dozen to hundreds (in various designs) of steam-driven pistons.
In 2018, the firm published papers on a spherical tokamak and a recent conceptual design was presented at the 30th IEEE Symposium of Fusion Engineering (SOFE). In August 2023, the company announced an updated plan to build a new fusion demonstration machine – Lawson Machine 26 (LM26) – at its Canadian headquarters. The company says LM26 is designed to achieve fusion conditions of over 100 million degrees Celsius (10 keV) by 2025 and progress towards scientific breakeven equivalent by 2026. This was an adjustment to its previously announced Fusion Demonstration Program. In June 2021, the company announced it would build 70% of a full-scale fusion demonstration plant in the UK as part of a public-private partnership with the UK Government.
History
The firm was founded in 2002 by former Creo senior physicist and principal engineer Michel Laberge.
In 2005 it produced a fusion reaction in its first MTF prototype. In 2010, it produced its first at-scale plasma injector with magnetically confined plasma. In 2011 it first demonstrated compressive heating of magnetized plasma.
A proof-of-concept compression system was built in 2013 with 14 full size pistons arranged around a 1-meter diameter spherical compression chamber to demonstrate pneumatic compression and collapse of a liquid metal vortex. The pneumatic pistons were used to create a converging spherical wave to compress the liquid metal. The 100 kg, 30 cm diameter hammer pistons were driven down a bore by compressed air. The hammer piston struck an anvil at the end of the bore, generating a large amplitude acoustic pulse that was transmitted to the liquid metal in the compression chamber. To create a spherical wave, the timing of these strikes had to be controlled to within 10 μs. The firm recorded sequences of consecutive shots with impact velocities of 50 m/s and timing synchronized within 2 μs. However it was found that the wall of the liquid metal vortex turned to a spray soon after the arrival of the pressure wave.
From its inception until 2016, the firm built more than a dozen plasma injectors. These include large two-stage injectors with formation and magnetic acceleration sections (dubbed "PI" experiments), and three generations of smaller, single-stage formation-only injectors (MRT, PROSPECTOR and SPECTOR).Peter O'Shea, Michel Laberge, Mike Donaldson, Michael Delage "Acoustically Driven Magnetized Target Fusion at General Fusion: An Overview" Poster presented at the 58th Annual Meeting of the APS Division of Plasma Physics 31 October – 4 November 2016. San Jose, California. CP10.00103 The firm published research demonstrating SPECTOR lifespans of up to 2 milliseconds and temperatures in excess of 400 eV.
As of 2016, the firm had developed the power plant's subsystems, including plasma injectors and compression driver technology. Patents were awarded in 2006 for a fusion energy reactor design, and enabling technologies such as plasma accelerators (2015), methods for creating liquid metal vortexes (2016) and lithium evaporators (2016).
In 2016 the GF design used compact toroid plasmas formed by a coaxial Marshal gun (a type of plasma railgun), with magnetic fields supported by internal plasma currents and eddy currents in the flux conserver wall.Russ Ivanov, Patrick Carle, Neil Carter, Ken Jensen, Stephen Howard, Michel Laberge, Alex Mossman, Peter O'Shea, Adrian Wong, William Young "SPECTOR 1 Plasma as a Target for Adiabatic Compression " Poster presented at the 58th Annual Meeting of the APS Division of Plasma Physics 31 October – 4 November 2016. San Jose, California. CP10.00106 In 2016, the firm reported plasma lifetimes up to 2 milliseconds and electron temperatures in excess of 400 eV (4,800,000 °C).
Around 2017 the company performed a series of experiments named Plasma Compression Small (PCS). These implosion experiments used a chemical driver to compress an aluminum liner onto a compact toroid plasma. Because the implosions involved chemical explosives, the tests took place outdoors in remote locations. The tests were destructive and could only be executed every few months. These tests were carried out to advance the understanding of plasma compression with the goal of advancing toward a nuclear-reactor scale demonstration.
, the PI3 plasma injector held the title as the world's most powerful plasma injector, ten times more powerful than its predecessor. It also achieved stable compression of plasma.
In 2019 it successfully confined plasma within its liquid metal cavity. From 2019 to 2021 it increased plasma performance.
As of 2021, the firm demonstrated compression of a water cavity into a controlled, symmetrical shape.
Also in 2021 the company agreed to build a demonstration plant in Oxfordshire, at Culham, the center of the UK's nuclear R&D. The plant was planned to be 70% of the size of a commercial power plant. The company claimed it had validated all the individual components for the demonstration reactor. However, as of 2026 the company's commercialization plan includes four years of system validation casting doubt on the earlier claims.
In 2022, the company announced that it had completed 200,000+ plasma shots, filed 150 patents/patents pending, and that headcount had passed 200. PI3 reached 10 ms confinement times and temperatures of 250 eV, almost 3 million degrees Celsius, without active magnetic stabilization, auxiliary heating, or a conventional divertor. Its primary compression testbed, a 1:10 scale system using water rather than liquid metal, has completed over 1,000 shots, behaving as predicted.
According to the 2023 Fusion Industry Association report, the company has 150 employees and has raised approximately US$300,000,000+.
In 2023, the firm reduced headcount significantly and announced that it was building a new machine, "LM26", with the goal of achieving breakeven by 2026. The Fusion Demonstration Plant being built in the UK will be delayed.
In May 2025, the firm again announced layoffs of about 25% of its workforce and reduction in operation of LM26, following an "unexpected and urgent financing constraint."
In January 2026, General Fusion announced plans to become a public company via a special purpose acquisition company merger by the middle of the year.
Organization General Fusion's CEO is Greg Twinney. The company's website states Greg joined General Fusion in 2020 with a well-established track record of executive leadership. Serving as General Fusion's chief financial officer for two years, he expanded the company's investor base and helped to launch the failed Fusion Demonstration Program. In 2022, he became General Fusion's chief executive officer.
Michel Laberge is the founder and Chief Science Officer of General Fusion. Prior to establishing General Fusion, Michel spent nine years at Creo Products in Vancouver as a senior physicist and principal engineer.
The board of directors is chaired by Klaas De Boer, who currently chairs AIM-listed Xeros Technology Group and serves on the Boards of SmartKem and veriNOS pharmaceuticals.
Technology
General Fusion's approach is based on the Linus concept developed by the United States Naval Research Laboratory (NRL) beginning in 1972. Researchers at NRL suggested an approach that retains many of the advantages of liner compression to achieve small-scale, high-energy-density fusion. According to Laberge, Linus could not properly time the compression using the technology of the era. Faster computers provide the required timing. However, this claim is not borne out by the literature as various Linus devices with no timing constraints, including systems using single pistons, were built during the experimental runs during the 1970s and demonstrated fully reversible compression strokes.
General Fusion's magnetized target fusion system uses a ~3 meter sphere filled with liquid metal. The liquid is spun, creating a vertical cavity in the centre of the sphere. This vortex flow is established and maintained by an external pumping system; liquid flows into the sphere through tangentially directed ports at the equator and exits radially through ports near the poles of the sphere.
A plasma injector is attached to the top of the sphere, from which a pulse of magnetically confined deuterium-tritium plasma fuel is injected into the center of the vortex. A few milligrams of gas are used per pulse. The gas is ionized by a bank of capacitors to form a spherical tokamak plasma (self-confined magnetized plasma rings) composed of the deuterium–tritium fuel.
The outside of the sphere is covered with steam pistons, which push the liquid metal and collapse the vortex, thereby compressing the plasma. The compression increases the density and temperature of the plasma to the range where the fuel atoms fuse, releasing energy in the form of fast neutrons and alpha particles.This energy heats the liquid metal, which is then pumped through a heat exchanger to generate electricity via a steam turbine. The plasma forming and compressing process repeats and the liquid metal is continuously pumped through the system. Some of the steam is recycled to power the pistons.
In addition to its role in compressing the plasma, the liquid metal liner shields the power plant structure from neutrons released by the deuterium-tritium fusion reaction, overcoming the problem of structural damage to plasma-facing materials. The lithium in the mixture breeds tritium.
Lawson Machine 26 (LM26) In August 2023, General Fusion announced it intends to build a new fusion demonstration machine named Lawson Machine 26 (LM26) to achieve important technical milestones using magnetized target fusion. LM26 aims to achieve fusion conditions of over 100 million degrees Celsius by 2025 and progress towards scientific breakeven equivalent by 2026.
LM26 will use a Marshall gun to inject a deuterium plasma into a target chamber. The target chamber's outer wall is a solid lithium liner contained within a cylindrical composite vacuum vessel. Toroidal coils mounted on the outside of the cylindrical vessel are pulsed and push on the liner to initiate compression. As the liner collapses, the plasma is compressed to higher density and temperature.
In January 2024, the company reported it had achieved symmetrical compression of a solid lithium ring within a few weeks of announcing LM26 and had built and began operating a compression test bed for LM26 named Prototype 0.
Fusion Demonstration Program The Fusion Demonstration Program is a 70% scale prototype which was being built in Oxfordshire, UK with a reported cost of US$400 million. It had been announced that the core technology had been proven out and was ready to be put together and that the plant was to commence operations in 2027. However the plant was put on hold in 2023 when the company announced that it would instead build a different machine in Canada aimed at demonstrating breakeven by 2026.
The plant had several key differences from the commercial power plant concept:
70% scale. 1 pulse per day repetition rate vs 1 pulse per second for a power plant. Among other things this increases the time available to re-establish the high vacuum conditions required for plasma formation by a factor of 86400, avoiding a significant engineering obstacle that will need to be solved for any future commercial application to be viable.