A chip fab needs more than 50 gas varieties to run its process steps, and China can make fewer than 30 percent of them. The two listed pure-plays in the business both grew their chip revenue about 29 percent in the first half of 2026, and one of them now supplies more than 90 percent of the country's 8-inch and 12-inch chip makers. The bottleneck is not chemistry. It is the certificate: a fab's approval takes 1 to 3 years, and a gas that fails it earns nothing.
What a fab actually buys
Electronic specialty gases are the high-purity gases that deposit films, etch patterns, dope silicon and clean chambers across the front-end process. Huate Gas calls them the second largest front-end semiconductor material after silicon wafers (2026 interim report, p. 13). Purity is the whole game: ultra-pure grades run to 4.5N, 5N, 6N and even 7N, meaning 99.99999 percent pure, with metal impurities, moisture and particles controlled to parts per billion (p. 15). A gas that fails purity does not just perform badly, it kills yield, so every product must pass a fab's certification before it earns a single order.
The market is growing fast and the gap is measured in varieties, not volume. SEMI data cited by Huate puts the global market above US$6bn in 2023, forecast to grow 8.5 percent a year to 2028, with China growing 12 percent, the fastest region (p. 14). At home, the specialty gas market grew from ¥17.5bn in 2017 to ¥49.6bn in 2023, an 18.96 percent annual rate (Jinhong Gas 2026 interim report, p. 16). The top end is still foreign: Linde, Air Liquide, Nippon Sanso, Air Products and Merck, which bought Versum's electronic gas business, have long dominated high-end semiconductor gases (Huate p. 16).
Two companies, two routes to the same fab door
Huate attacks from catalogue breadth. The company says 57 electronic specialty gas products have achieved import substitution, more than 20 are already used on 14nm and 7nm lines, and some fluorocarbon and hydride products have entered 5nm processes (Huate p. 16). Its products serve more than 90 percent of domestic 8-inch and 12-inch chip makers (pp. 16, 24). Three newer products, hydrogen bromide, boron trichloride and sulfur tetrafluoride, passed certification at some semiconductor makers in the half (p. 26), and a 1,936.2 tonne per year electronic specialty gas project was completed in August 2026 (p. 25). The business is growing: semiconductor-field revenue was ¥342.89m in the half, up 28.22 percent, inside total revenue of ¥872.39m, up 28.95 percent (pp. 7, 24).
Jinhong Gas attacks from scale and service. It is the largest private gas supplier in China by sales, sells more than 100 gas varieties, and runs a total gas and chemical management service that handles a fab's whole gas supply, from cylinders and piping to analysis and waste handling (2026 interim report, pp. 18, 19, 23). It holds 418 patents, 124 of them invention patents, and runs the gas industry's only national enterprise technology centre focused on electronic gases (pp. 23, 28). Its integrated circuit revenue rose 29.56 percent in the half, and it says it is deepening strategic cooperation with SMIC, CXMT and SK Hynix (p. 25). But scale has a cost: revenue rose 12.12 percent to ¥1,473.21m while net profit fell 17.41 percent to ¥67.89m, and core net profit excluding non-recurring items fell 73.50 percent to ¥17.82m as new projects built against the cycle started depreciating (pp. 11, 12).
The helium windfall flatters the numbers
Both companies' halves were flattered by helium, not chip gases. Geopolitical disruption pushed helium prices up, and Huate's helium and related revenue, about 20 percent of its total, grew 133 percent in the half (Huate p. 24). Jinhong's helium revenue grew 140.53 percent, with second-quarter helium revenue up 110.01 percent on the first quarter (Jinhong p. 26). That windfall is fragile. Huate says its helium supply depends on Russian export permits that must be applied for monthly, with uncertain outcomes (Huate p. 43), and Jinhong's filing is equally clear that helium depends on geopolitically disrupted supply, so its 140 percent growth rate is not a run rate to extrapolate. Strip helium out and the chip-gas story underneath is steadier but smaller: both companies grew semiconductor revenue about 29 percent in the half, on new fab capacity coming online at home.
Our view
The gas business is the least glamorous bottleneck in the chip chain and one of the most honest. The chemistry is not the hard part, qualification is, and the filings show both companies still earning their advanced-node credentials one fab at a time. Huate has the better technology story, with products already introduced into 5nm processes, and Jinhong the better scale and service story, but the honest measure is the same for both: repeat orders at advanced fabs, not growth percentages off a helium-inflated base. Until a Chinese gas maker can point to repeat volume at a leading-edge fab, the sub-30 percent figure is the story, and the 29 percent growth is the footnote.
What to watch
- Whether Huate's 5nm introductions turn into repeat volume at leading-edge fabs, rather than staying as sample qualifications.
- Whether Jinhong's new capacity fills fast enough to cover the depreciation that is currently eating its profit.
- Whether helium stays a windfall or reverts: if prices fall back, both companies lose the cushion that flattered this half's results.
- New certifications: hydrogen bromide, boron trichloride and sulfur tetrafluoride passed at some makers in the half; watch for named fabs and repeat orders.
The full picture is on our semiconductors industry map, with each stage's companies, volumes and margins.