For a decade, Beijing's answer to the chip squeeze was a chequebook with a name. The third and largest version of that chequebook has stopped sitting idle — and the first places it writes to say more than the headline number.
The National Integrated Circuit Industry Investment Fund Phase III (国家集成电路产业投资基金三期), widely called "Big Fund III," is the clearest signal yet of where China thinks its semiconductor weakness still lives.
What 'Big Fund III' actually is
Big Fund III is a state-backed investment vehicle, not a government department. It was registered on 24 May 2024 with a mandate to invest in the semiconductor supply chain through market-style vehicles. It follows two earlier funds: Phase I, launched in September 2014 at about ¥138.7B, focused on building fabrication capacity; Phase II, launched in October 2019 at over ¥200B, leaned toward equipment and materials.
Phase III is the largest by a wide margin — its registered capital alone exceeds the sum of the first two.
The numbers that matter
- Registered capital: ¥344B (≈ US$48.5B / HK$378B).
- First deployment: on 31 December 2024 it seeded two sub-funds — Huaxin Dingxin (华芯鼎新) at ¥93.1B and Guotou Jixin (国投集新) at ¥71.1B, a combined ¥164B (≈ US$23.1B / HK$180B).
- Bank commitment: the six major state banks pledged ¥114B (≈ US$16.1B / HK$125B), about 33% of the fund, to be paid in over ten years from establishment.
These figures are drawn from company registration records and Chinese state financial media, not from speculation.
Who is putting up the cash
The fund has 19 shareholders. The Ministry of Finance and CDB Financial sit at the top, joined by Shanghai Guosheng and the six big state banks — ICBC, CCB, ABC, BOC, BOCOM and PSBC.
The bank breakdown is instructive:
- ICBC, ABC, BOC and CCB each committed ¥21.5B (a 6.25% stake).
- BOCOM committed ¥20B (5.81%).
- PSBC committed ¥8B (2.33%).
This is the first time the large state banks have appeared as shareholders in the Big Fund structure, and it tells you the effort is now treated as strategic infrastructure, not a side bet.
Where the first cheques went
The two launch sub-funds split the work. Reporting on the fund's stated allocation says roughly 70% is aimed at the upstream — equipment, materials and core components — with about 30% directed at newer fronts like AI chips, advanced packaging and high-end storage.
Concrete deployments have started to surface. In 2026, Guotou Jixin took a roughly ¥950M (≈ US$134M / HK$1.04B) stake in Primarius (概伦电子), a domestic EDA software firm, and the fund has been linked to equipment and photomask-substrate projects. A separate, larger commitment backs Hua Hong's FAB9B project, reported at about US$4.17B, where equipment purchasing is expected to pull domestic suppliers into the order book.
These are reported picks rather than a full public ledger, so treat the specifics as directional.
Why the aim moved upstream
The shift from "build fabs" to "fix the tools and materials" is the whole point of Phase III. China has made real progress in mature-node manufacturing, but the hardest, most import-dependent links are still the ones few talk about: lithography-adjacent equipment, photomask substrates, specialty gases, and the EDA software that designs cannot proceed without.
The rationale is blunt. China's integrated-circuit imports still ran at about US$361.8B in the first seven months of 2026, up 58.3% year on year, with a trade gap near US$145.8B, according to the National Development and Reform Commission. A country that assembles the world's electronics cannot treat those inputs as someone else's problem.
What it does not fix
Two things the chequebook cannot buy.
First, talent. Industry bodies estimate China's semiconductor talent shortfall at more than 300,000 people across 2025–2026, widening toward 500,000 by 2030. Money trains fewer masters in a hurry.
Second, the frontier. The fund can strengthen the domestic base in equipment and materials, but the very top nodes — sub-7nm logic without restricted lithography — remain a physics and tooling problem that capital alone does not close.
Honest limitations
This article relies on Chinese state media, company registration filings and analyst commentary; we did not access the fund's internal investment committee documents. The 70/30 allocation split is reported guidance from analysts and fund-tracking coverage, not a binding public mandate. Specific 2026 deployment names (Primarius stake, Hua Hong FAB9B, photomask-substrate projects) are reported by Chinese financial media and should be read as indicative, since the fund does not publish a live portfolio. Currency conversions use fixed illustrative rates (US$1 ≈ ¥7.1, HK$1 ≈ ¥0.91) and will drift with markets. None of this is investment advice.
What readers can do now
- Watch the upstream, not the headline ¥344B: equipment, materials and EDA suppliers are where the money is actually pointed.
- Track whether Phase III's sub-funds report real, recurring orders for domestic toolmakers — that is the only proof the strategy is working.
- Note the policy backdrop: advanced-node projects under 28nm with long operating lives keep generous tax breaks, which changes the cash maths for funded firms.
- For observers outside China, treat Big Fund III as a structural signal about where industrial policy is headed, not as a stock screen or a forecast of technical breakthroughs.