BEST Fusion Coil Delivery: China's Schedule Edge Over CFS's SPARC
Sinomach has shipped a batch of eight toroidal-field coil cases for China's BEST tokamak in 14 months, keeping the machine on track for completion at the end of 2027. Commonwealth Fusion, the venture-backed leader in the West, has installed eight of the 18 magnets its SPARC machine needs and has already slipped its own target to 2027. The desk reads the gap as schedule discipline and prices it through the component makers landing repeat orders.
Fusion news this week came down to a shipping date. On Sunday, the state-owned heavy-industry group Sinomach Heavy Equipment marked the completion of a batch of eight toroidal-field coil cases for China's BEST tokamak, built by its China Erzhong unit at Deyang. Commonwealth Fusion Systems, the venture-backed leader of the Western field, said this month that it has installed eight of the 18 toroidal-field magnets its SPARC machine needs. Both programmes are at the eight-unit mark, but the milestones differ: Sinomach has completed a production batch while SPARC is still installing magnets. China is showing it can turn fusion into a manufacturing schedule; Western firms are still proving the physics.
Sichuan Online's report on the Deyang ceremony describes BEST, the Burning plasma Experimental Superconducting Tokamak, as designed by the Institute of Plasma Physics under the Chinese Academy of Sciences and built by Fusion Energy Tech Anhui at Hefei's Future Science City. Construction formally started in January 2023, the target completion date is the end of 2027, and Sinomach's coil-case batch took 14 months to produce.
The design calls for deuterium-tritium burning plasma, a fusion power output of 20 to 200 megawatts, and a Q greater than one, meaning the plasma releases more fusion energy than is spent heating it. A demonstration of actual fusion electricity is pencilled in for around 2030. The coil cases delivered this week, along with a magnet component that passed its maximum current test in March, are the physical evidence behind that target.
Sinomach's role and delivery speed make the batch notable. Sinomach's China Erzhong unit had its first coil case accepted in March, delivered the rest of the batch in the following months, and has already signed a follow-on contract with the Institute of Plasma Physics for a batch of shielding blocks, the profile of a factory working a repeat order book. IT Home's coverage situates the delivery inside China's wider tokamak push: EAST ran a 1,066-second high-confinement plasma pulse above 100 million degrees Celsius in January, a duration record, and the Chengdu-based HL-3 machine is aiming at its own burning-plasma experiments around 2027.
For investors, the opportunity lies in BEST's supply chain. Hefei Metalforming, listed in Shanghai, is building the machine's eight-sector vacuum vessel, and Invengo's Rongke Hengyang subsidiary supplies fusion power systems. Sinomach's heavy-industry unit now has two consecutive contracts, coil cases and shielding blocks, from the same customer, which is what a repeat vendor with a growing order book looks like. China is aiming for commercial fusion power by 2040, and Cailianshe's report from last week's World Manufacturing Convention in Hefei shows how eagerly listed suppliers now present themselves as part of that plan. The A-share "fusion concept" basket trades on the 2040 horizon, well ahead of any revenue tied to it. Component makers with delivered, accepted hardware and follow-on orders are a different investment case from a ticker that only carries the theme.
Commonwealth Fusion Systems' SPARC programme offers a comparison. SPARC is privately funded and moves on venture timelines that have already slipped: the company said in 2024 that it expected first plasma by the end of 2026, and that target is now 2027. Its first toroidal-field magnet was installed at the start of this year, roughly the same window in which Sinomach began delivering coil cases for BEST. One programme is backed by a state contractor with a production schedule and a follow-on order book. The other depends on continued private capital reaching milestones on a compressed timeline. That gap is about execution risk, and it makes China's schedule the easier one for an investor to price with any confidence.
On-time coil deliveries demonstrate manufacturing execution, not the physics. A 2027 assembly date for BEST is still far from demonstrated burning plasma. An engineer at one of the component suppliers put the optimistic version of the counter-case bluntly: fusion has been "50 years away" for decades, and what is left are engineering problems, which the argument holds are China's comparative advantage. That line is closer to marketing than to physics. Reaching Q greater than one on a deuterium-tritium plasma is a plasma-physics and materials problem that a well-run supply chain cannot solve by itself, and a finished tokamak shell can sit for years while operators tune confinement.
The desk accepts the physics risk but considers the schedule evidence important. What a state-backed, batch-production supply chain buys is certainty about the parts of fusion that are genuinely engineering: coil cases, vacuum vessels, shielding blocks, power supplies. When a burning-plasma result does or does not arrive, it will not be because the hardware showed up late. The Deyang ceremony supports the narrower manufacturing claim; "China wins fusion," by contrast, goes beyond what it demonstrates. It moves the investable risk away from the tokamak programme itself and toward the specific vendors who keep landing repeat, accepted orders.
The desk expects BEST's main assembly to complete in 2028, a modest slip against the stated end-2027 target and typical of a first-of-a-kind superconducting device, with first deuterium-tritium burning plasma no earlier than 2029. Before either milestone, watch for further batch contracts, on shielding and divertor components in particular, going to the same tier of state-owned suppliers already building the coil cases and vacuum vessel. Each contract would indicate that the schedule is holding and could move the A-share fusion basket, whether or not the underlying plasma physics advances.