The key race: SPARC's magnet campaign runs into September, ITER keeps beating its own schedule, and China tests the largest fusion magnet ever built
Commonwealth Fusion Systems' SPARC in Devens, Massachusetts was still installing toroidal-field magnets as of late August 2026, with the company's own 'end of summer' target for finishing all 18 now bumping against the calendar and first plasma plus a Q greater than 1 net-energy attempt still held for 2027. The commercial side of the business moved on a longer clock this cycle: having filed the industry's first-ever grid-interconnection request for a fusion plant with PJM in April 2026, CFS's own published timeline for the Fall Line Fusion Power Station in Chesterfield County, Virginia now reflects PJM's actual four-to-six-year study process, stretching the plant's 'in service' window from an earlier 2026-2028 estimate to 2026-2032, even though the harder local hurdle, Chesterfield County's zoning and conditional-use approval, was already secured back in September 2025.
The 34-nation ITER project kept running ahead of its own baseline rather than behind it: after lowering the sixth of nine tokamak sector modules into the machine's assembly pit on July 28, almost six months ahead of the original plan and putting two-thirds of the torus-shaped core in place, ITER now expects to lower a seventh sector module before the end of 2026 while still targeting mid-2027 for the ninth and final one. China answered with its own hardware milestone rather than a schedule claim: in June 2026 its CRAFT engineering platform in Hefei finished testing a 582-tonne D-shaped toroidal-field magnet, the largest superconducting magnet ever built for a fusion reactor, rated to 46.5 kiloamps and tested past 60. It is one of 16 such coils being built for the BEST tokamak, which remains on track to complete assembly around 2027 on the way to a stated 2030 power-generation demonstration.
Who's ahead, and on which path: General Fusion answers to public shareholders for the first time, Pacific Fusion breaks ground, and a fourth laser entrant solves its fuel problem
General Fusion gave the sector its first real taste of public-market scrutiny: after becoming the first fusion company to trade publicly on July 13, it filed its first quarterly business update as a public company on August 18, disclosing roughly $150 million in cash following its SPAC merger and reporting progress on LM26, its magnetized-target-fusion machine in Vancouver, toward a 1 kilo-electron-volt electron-temperature milestone. The stock, trading on Nasdaq as GFUZ, has swung hard since debut and sat around $7.16 on August 28, evidence investors are still working out how to price a pre-revenue fusion company now that one is actually public and reporting quarterly. Pacific Fusion converted planning into steel the same week: it broke ground on its $1 billion Mesa del Sol research and manufacturing campus in Albuquerque on August 25, the site meant to house the 156-module Demonstration System targeting net facility gain by 2030, and signed a fusion-research memorandum of understanding with the National Nuclear Security Administration the same day, with two of the more than 150 full-scale pulser modules already completed at its nearby Los Lunas build center.
On the inertial side, Inertia Enterprises, the Livermore, California laser-fusion startup founded by Twilio's Jeff Lawson alongside NIF veterans Annie Kritcher and Mike Dunne, said in mid-August that it had found a way to manufacture its fusion fuel targets quickly, the unglamorous but load-bearing problem every laser-fusion company has to solve before it can fire at any commercially relevant repetition rate; the company opened a 50,000-square-foot Livermore headquarters in July to house both that target factory and Thunderwall, its planned high-energy laser. Fuse Energy, fresh off its record 1.27-trillion-neutron shot in mid-August, deepened its government ties rather than chasing a second record, signing a five-year cooperative research agreement with the Nevada National Security Site's Mission Support and Test Services on August 2 covering tritium capability and pulsed-power industrial-base work, on top of existing agreements with Los Alamos and Sandia. TAE Technologies added a supply-chain hedge of its own, signing an August 5 helium-3 fuel-supply agreement with lunar-resource startup Black Moon Energy, while Trump Media & Technology Group's August 10 Form 8-K reaffirmed the two companies' roughly $6 billion merger is still targeted to close by the end of 2026, even though the Form S-4 needed to register the deal had still not been filed. Tokamak Energy, mid-upgrade on ST40, used the quiet window to publish an 11-paper Focus Collection in the journal Nuclear Fusion on August 18, covering the plasma physics, engineering systems and commercial case for a fusion pilot plant.
What decides it: whether this year's magnet, coil and neutron records convert into machines that actually run, and whether Washington's own de-risking model keeps pace
The physics gates are unchanged from earlier this year: whether SPARC crosses Q greater than 1 in 2027, whether China's BEST tokamak reaches first plasma on its own 2027 timeline, and whether pulsed and inertial approaches convert single-shot records, Fuse's neutron yield, Pacific Fusion's 440-gigawatt pulser test, into repeatable, power-plant-relevant performance rather than one good shot. General Fusion's first quarterly report as a public company is now a live test of a different kind: whether a market that has never had to price a fusion company's technical risk in real time settles into something more stable than the roughly $6-to-$15 swing its stock has shown since debut, a question every other fusion company eyeing a public listing, TAE included, is watching closely.
That test is unfolding against public money that has not kept pace with the private capital it is meant to unlock. The Department of Energy's Milestone-Based Fusion Development Program, the model credited with turning $46 million of federal seed funding into more than $350 million of private capital across eight companies, remains authorized for $415 million through fiscal year 2027 under the CHIPS and Science Act but has received only about $98 million in actual appropriations to date, a gap the Fusion Industry Association has been pressing Congress to close. Nearly every company in this sector still depends on a mix of private capital, corporate offtake, and public co-funding to bridge the years between a plasma record and a power plant, and that shortfall is the clearest test of whether Washington's own pitch, that a small federal dollar reliably pulls in many private ones, keeps working as the sector matures past its earliest movers.
The money and the rules: no new funding record since July, but a regulator still racing its own October deadline
No new sector-wide funding tally has landed since the Fusion Industry Association's 2026 Global Fusion Industry Report on July 13, which counted a record $4.48 billion raised in the twelve months to July, up 69 percent year over year across 56 surveyed companies, taking cumulative sector funding to $14.24 billion. What has changed since is composition rather than scale: General Fusion's Nasdaq listing turned a slice of that private capital into a publicly traded, dollar-priced stock for the first time, Pacific Fusion converted committed capital into a physical campus with its Albuquerque groundbreaking, and CFS, Tokamak Energy and Proxima Fusion are all now generating supplementary revenue through magnet sales, defense contracts and government partnerships rather than equity alone. China's parallel state commitment keeps compounding regardless: fusion remains one of eight frontline technologies named in the 15th Five-Year Plan (2026-2030), and the CRAFT platform's 582-tonne magnet is the clearest recent proof point that Beijing's stated intent to out-invest the sector is showing up in finished, tested hardware rather than announcements alone.
On regulation, the US Nuclear Regulatory Commission remains on track to finalize its 'Regulatory Framework for Fusion Machines' as early as October 2026, having closed public comment on the proposed rule on May 27, well ahead of the December 31, 2027 statutory deadline; the rule would keep fusion regulated under lighter byproduct-materials rules rather than as a fission reactor, the approach Washington State already used to issue Helion the world's first licenses to build and operate a fusion plant. No text of a final rule had been published as of this update, so the sector's regulatory posture is unchanged from earlier this year, just closer to the deadline that matters.
What to watch through 2027-2028
At Commonwealth Fusion, watch whether the remaining SPARC toroidal-field magnets finish installing and then the defining event, first plasma and a Q greater than 1 attempt in 2027; watch the Fall Line Fusion Power Station's PJM interconnection study now that it is running on a realistic multi-year clock rather than an optimistic one. At ITER, watch whether the seventh sector module goes in before the end of 2026 and whether the mid-2027 target for the ninth and final module holds; watch China's BEST tokamak convert its 582-tonne magnet and the rest of its coil set into a completed machine around 2027. At Tokamak Energy, watch the DOE/DESNZ-funded ST40 upgrade wrap toward a 2027-2028 restart, and whether UK-based HTS tape manufacturing with Furukawa moves from evaluation to commitment. At Proxima Fusion, watch completion of the Stellarator Model Coil in 2027, the hardware gate before Alpha construction begins near Munich.
At Pacific Fusion, watch the first full-scale production pulser module clear testing and watch progress toward the 2027 detailed-proposal call for its Users Program; at Helion, watch Tiny Merge come fully online and start feeding design choices into Orion ahead of the 2028 Microsoft delivery deadline. At TAE, watch whether the roughly $6 billion TMTG merger actually closes by the end of 2026 now that the Form S-4 is the last disclosed procedural gap, and whether a Da Vinci site gets named from the Alabama, Ohio and Texas shortlist. On the newer entrants, watch whether General Fusion's stock finds a floor as more quarterly results land, whether Inertia Enterprises' fuel-target breakthrough translates into an actual Thunderwall test campaign, and whether Fuse Energy converts its NNSS agreement into recurring government revenue. On policy, watch whether the NRC finalizes its fusion rule as targeted in October 2026, and whether Congress closes the gap between the Milestone-Based Fusion Development Program's $415 million authorization and its roughly $98 million in actual appropriations, still the clearest test of whether Washington's own de-risking model keeps pace with the industry it helped create.
Sources
- Fusion Industry Association - 2026 Global Fusion Industry Report ($4.48B annual, $14.24B total)
- ITER Organization - Two-thirds of ITER tokamak core now in place
- US NRC - Fusion Machine Rulemaking Status
- MIT News - MIT researchers tackle the economic realities of fusion power (Economic Q framework)
- GlobeNewswire - General Fusion provides first public company business update
- PR Newswire - Fuse deepens partnership with Nevada National Security Sites (CRADA)
