Other paths to fusion: mirrors, FRCs, pinches and spheromaks

A magnetic mirror confines plasma in a straight tube instead of a closed ring: coils at the ends of the tube raise the magnetic field there, and that gradient reflects most of the charged particles approaching it back toward the center, like a ball bouncing off a slope. The classic problem with mirrors, studied since the 1950s and 60s, is that some particles fall inside a velocity-space "loss cone" and escape through the ends anyway, which historically gave them worse confinement than a tokamak. Interest was renewed by high-temperature superconducting tape magnets, which allow much higher fields in smaller, cheaper mirrors. WHAM (Wisconsin HTS Axisymmetric Mirror), a collaboration between the University of Wisconsin-Madison, the company Realta Fusion and Commonwealth Fusion Systems — which built the magnets —, began plasma operations in July 2024 in Wisconsin with magnets of up to 17 T, according to those institutions; its goal is to study whether a high-field axisymmetric mirror can serve as a neutron source and, later, a reactor.

A field-reversed configuration (FRC) is a donut-shaped plasma that generates its own poloidal field with an internal current, needing neither a surrounding toroidal chamber nor external coils threading the central hole: in that sense it is the most compact magnetic design per unit of applied field. It operates at a beta close to 1 — plasma pressure comparable to the magnetic pressure confining it — which in theory allows smaller reactors per unit of applied field. The company TAE Technologies, in California, operates Norman (in operation since 2017), which sustains the plasma by injecting neutral particle beams, and is building Copernicus, its next machine, per the company. TAE reported in 2021 having sustained plasmas above 50 million degrees in Norman and, in later statements, higher figures; the machine's results are described in the peer-reviewed literature, while the more recent figures the company circulates still have no associated publication. TAE is aiming for an aneutronic fuel, proton-boron-11, which in theory nearly eliminates neutron production, though it requires much higher temperatures than deuterium-tritium.

A Z-pinch compresses a plasma column by running a very high electric current along its axis: the magnetic field that current itself generates squeezes it inward. That is the principle behind the Z Machine at Sandia National Laboratories, in New Mexico, the world's most powerful pulsed-power generator, used since 2013 for magnetized liner inertial fusion (MagLIF) experiments: it compresses, with up to 20 million amperes, a laser-preheated, pre-magnetized beryllium tube, and Sandia has reported yields of up to 1.1 × 10¹³ deuterium-deuterium fusion neutrons per shot. The company Zap Energy, spun out of University of Washington research, pursues a variant called the sheared-flow-stabilized Z-pinch, which uses the plasma's own motion to keep the column from kinking and collapsing, with no need for external fields or lasers. In 2022 LLNL reported, together with the University of Washington, peer-reviewed measurements (Physics of Plasmas, 2021) confirming thermonuclear fusion neutrons in the FuZE device, of which Zap Energy is the corporate spinoff.

A spheromak, like an FRC, is a plasma sustained by its own internal current with no external toroidal coils, but with a different magnetic topology, closer to that of an apple than a donut. It was studied mainly in the 1980s and 90s, among other places in the Sustained Spheromak Physics Experiment (SSPX) at Lawrence Livermore National Laboratory, which between 1999 and the mid-2000s sustained toroidal currents of up to about 600 kA and electron temperatures of 0.5 keV for a few milliseconds. Most groups that studied it have since moved on to FRCs or mirrors.

Magnetized target fusion combines a pre-magnetized plasma, as in an FRC or a spheromak, with a mechanical or current-pulse compression that pushes it to fusion conditions within microseconds, halfway between magnetic and inertial confinement. Besides Sandia's MagLIF, the Canadian company General Fusion is pursuing this path with its Lawson Machine 26 (LM26) demonstration device: it injects a spherical-tokamak deuterium plasma and compresses it with an imploding solid lithium liner; the company's commercial plant concept, unlike LM26, uses pistons that compress a liquid lithium liner. In June 2026 General Fusion reported compressional heating to about 0.72 keV (about 8.4 million degrees) of electron temperature in LM26, in a technical paper submitted for peer review but not yet published. The temperature and performance figures companies release on their own, without an accompanying peer-reviewed publication, should be read for what they are: a company's own claims about its own equipment, still without external verification.