China's Chip Stack: Export Controls, Huawei and the Race for Self-Reliance
Since 2022 the United States has tried to cut China off from the most advanced AI chips, the machines that make them and the memory they need. China's answer has been to build a second, parallel AI stack: Huawei's Ascend accelerators made at SMIC without EUV lithography, home-grown HBM, a new software ecosystem, and labs such as DeepSeek that squeeze more out of less. This guide maps that stack layer by layer, measures where it is strong and where it is stuck, and asks whether the controls are slowing China down or speeding up its independence.
๐งฑ Two AI stacks
Every earlier part of this series described one global supply chain: Nvidia designs, TSMC fabricates with ASML machines, SK hynix supplies memory. Since October 2022, US export controls have tried to keep China out of the top of that chain. China cannot legally buy Nvidia's Blackwell or Rubin chips, ASML's EUV machines or the latest HBM, and most of its access to TSMC's leading nodes is closed.
Yet Chinese labs keep shipping competitive models, often open-weight, and Huawei has published a roadmap of a new Ascend chip every year to 2028. The result is a second AI stack, less efficient and more expensive at almost every layer, but increasingly complete.
๐ฐ๏ธ Controls and countermoves
โ๏ธ Two-lane timeline
US actions on the left, China's responses on the right. Filter by lane.
๐บ๏ธ The Chinese stack, layer by layer
๐งฉ How self-sufficient is each layer?
Click a layer. Ratings are our qualitative summary of the reporting, as of October 2026.
๐ญ SMIC: advanced chips without EUV
SMIC, China's largest foundry, cannot buy EUV machines. To print features small enough for 7 nm and 5 nm-class chips, it uses older deep-ultraviolet (DUV) immersion scanners and multi-patterning: exposing each critical layer two, three or four times with shifted masks. It works, as Huawei's phone and Ascend chips show, but every extra exposure adds steps, time and chances for defects.
2025 reports put SMIC's 5 nm-class yields at roughly a third of TSMC's on comparable chips, with wafer costs perhaps 50% higher. Epoch AI says the Ascend 950 and 960 compute dies use SMIC's N+3 process (a 5โ7 nm-class DUV node), with N+4 expected for the 970, though other reports describe N+3 as still maturing because of a shortage of advanced DUV scanners. Reliable figures for SMIC's advanced-node capacity are not public.
๐งฎ What multi-patterning costs
๐๏ธ Cost per working AI die: EUV vs DUV multi-patterning
Standard yield model (good dies = eโarea ร defect density). All inputs illustrative; defaults loosely follow the 2025 reports.
Dies per 300 mm wafer approximated as (ฯยท150ยฒ รท area) โ (ฯยท300 รท โ(2ยทarea)). Real yields also depend on redundancy, binning and packaging losses.
๐ด Huawei Ascend: the national champion
At its September 2025 Connect conference, Huawei published its first multi-year chip roadmap: the Ascend 950PR (Q1 2026, tuned for inference prefill and recommendation), 950DT (Q4 2026, decode and training), 960 (2027) and 970 (2028), each roughly doubling compute, and from the 950 onwards using Huawei's own HBM. Earlier Ascend 910B and 910C chips were made on SMIC's 7 nm process; teardowns of some 910-series parts found TSMC-made dies obtained through intermediaries before TSMC cut off supply.
Huawei's strategy is to win at the system level. If each chip is weaker, connect more of them, with a powerful network. The Atlas 950 SuperPoD joins 8,192 Ascend chips into one machine, versus 72 GPUs in an Nvidia NVL72 rack. Epoch AI estimates median Ascend output of about 1.5 million units in 2026.
โ๏ธ Chip vs system: closing the gap with scale
Vendor-announced peak FP8 compute (dense). Toggle between per-chip and per-system. Different precision definitions make exact comparisons uncertain.
Ascend 950: ~1 PFLOPS FP8 per chip; Atlas 950 SuperPoD: ~8 EFLOPS FP8 (Huawei, Sep 2025). Nvidia B200: ~4.5 PFLOPS dense FP8; GB200 NVL72: ~360 PFLOPS dense FP8 (Nvidia specifications). Power and space are not shown: the SuperPoD needs far more of both.
๐พ Memory: the binding limit
Analysts increasingly argue that HBM, not logic chips, sets how many accelerators China can build. US controls since December 2024 restrict HBM sales to China. Chinese firms are reported to have stockpiled about 11.4 million HBM stacks from Samsung before the rules tightened, but stockpiles run down.
China's answer is CXMT, its leading DRAM maker. Reports say it aims for HBM3 mass production by the end of 2026, with analysts warning of slippage; an April 2026 report said no mass-production orders had yet been placed. In August 2026 The Information reported that CXMT had begun small-scale HBM3E production, which the company has not confirmed. Huawei says its own HBM will appear in the Ascend 950. Even on the optimistic reading, Chinese HBM is a generation or two behind SK hynix's HBM4 (see The Memory Wall).
๐ The challengers and the money
๐ China's AI chip companies
| Company | Main products | Latest numbers | Notes |
|---|---|---|---|
| Huawei (HiSilicon) | Ascend 910C, 950PR/DT; Atlas SuperPoDs; CANN software | ~1.5M Ascend units in 2026 (Epoch median estimate) | Vertically integrated: chips, HBM, systems, cloud, models |
| Cambricon | Siyuan (MLU) accelerators | Revenue ยฅ9.53B in JanโSep 2026, +107%; Q3 ยฅ3.53B, net profit ยฅ1.40B | Profitable; growth slowing sequentially; supply capacity a constraint |
| Moore Threads | MTT GPUs | IPO Dec 2025 on Shanghai's STAR Market: raised ยฅ8B; shares +468% on day one | Loss-making |
| MetaX | GPUs for AI and HPC | IPO Dec 2025: ~$600M raised; shares +~700% on day one | 2024 revenue ยฅ743M, net loss ยฅ1.4B |
| Hyperscaler chips | Alibaba (T-Head), Baidu (Kunlun), others | Limited public data | Mostly for their own clouds |
| CXMT | DRAM, HBM3 (planned) | H1 2026 revenue ยฅ150B (+874%) amid the memory boom | HBM3 mass production targeted late 2026 |
Cambricon Q3 report (23 Oct 2026, via TechFlow); Caixin; DigiTimes; CNBC; Taipei Times; TrendForce; roic.ai/The Information. Some figures from single reports.
Software is the quieter battle. Huawei's CANN stack and the MindSpore framework are Huawei's answer to CUDA, and Chinese labs increasingly port models to Ascend. Developers report the tooling is less mature, which adds engineering cost to every model.
๐ฆ The last wall: lithography
The deepest gap is the one ASML fills in The Chokepoint. Reuters reported in December 2025 that a Chinese EUV prototype had been completed in early 2025 in Shenzhen, built with the help of former ASML engineers, and was being tested. By August 2026 reports said it generates 13.5 nm EUV light but has not yet produced a working chip; expectations for chip production range from 2028 to 2030. Domestic DUV scanner makers are also progressing, which matters because SMIC's multi-patterning depends on DUV tools that are themselves export-controlled.
๐ณ๏ธ Leaks and workarounds
- Smuggling
- In October 2026 US prosecutors arrested a California man accused of shipping servers with about $300 million of Nvidia AI chips to China through US freight companies and Malaysian transshipment firms. Taiwan opened an investigation in May 2026 into servers allegedly exported with forged documents, and in December 2025 the owner of Hao Global pleaded guilty in another smuggling case. Bloomberg reported in October 2026 that officials say Nvidia is not doing enough; Nvidia says it follows the rules.
- Licensed sales
- The H20, designed for China, was halted in April 2025 and later licensed; Nvidia's 10-K says US officials "expressed an expectation" of 15% of the revenue, never codified. The H200 was cleared in December 2025, and a BIS rule effective 15 January 2026 moved H200- and MI325X-class chips to case-by-case review. About ten Chinese firms were approved, but a Commerce official told Congress in July 2026 that shipments remained "very few"; Nvidia's first licensed H200 sales came in the quarter to 26 July 2026 and were under 1% of its $89 billion data-centre revenue. Sources describe the 25% charge differently: as a revenue share, or as a Section 232 tariff on chips imported into the US for re-export.
- China's own brakes
- Beijing has reportedly told domestic firms to stop buying Nvidia chips (from September 2025) and to favour Huawei, which helps explain why licensed sales are small.
- Remote access
- Chinese companies can rent GPUs in data centres outside China, a loophole regulators have debated closing.
- Efficiency
- DeepSeek trained V3 on Nvidia's H800, a cut-down China-market chip, and published unusually detailed papers on squeezing performance from limited hardware (see its hardware reflections paper). Constraint has pushed Chinese labs to excel at efficiency.
โ๏ธ Are the controls working?
๐บ๐ธ The case that they work
China's total AI compute is far smaller than America's, and each unit costs more. SMIC's yields and HBM shortages cap volumes. Frontier training runs need hundreds of thousands of top chips, which China cannot assemble at the same scale. Every Chinese chip is a generation or more behind and uses more power for the same work.๐จ๐ณ The case that they backfire
Nvidia's China share has collapsed and Huawei, Cambricon and others have gained a protected home market, funding rapid iteration. Chinese labs keep releasing strong open models. Controls created urgency and state funding that may produce a durable rival stack, including EUV, within a decade.Both can be true: the controls slow China's total compute in the near term while accelerating its long-term independence. Which effect dominates depends on how fast China closes the HBM and lithography gaps.
๐ฎ Scenarios to 2030
๐ฒ Explore the futures
Pick assumptions; the panel summarises a plausible outcome. Qualitative and illustrative, not a forecast.
๐ญ What to watch
- Ascend 950DT and Atlas 950 SuperPoD deliveries in Q4 2026, and real-world training results on them.
- CXMT HBM3 mass production and Huawei's own HBM: the clearest signal of whether China can scale accelerator output.
- SMIC N+3 yields and capacity, and any new restrictions on DUV tools and servicing.
- US licensing: whether H200 shipments grow, whether Blackwell-class chips are ever licensed, and any rules on remote cloud access.
- The EUV prototype: any first working chip would be a major milestone.
- Chinese frontier models trained mainly on domestic chips.
๐ข Reading the numbers
- Huawei output (~1.5M) is Epoch AI's median estimate; Chinese companies do not publish volumes.
- Chip specifications are vendor-announced peaks; Huawei and Nvidia define precisions differently, so the comparison is approximate.
- SMIC yields and costs come from 2025 reports and may be outdated.
- HBM progress (CXMT HBM3E) is reported but unconfirmed; the 11.4M stockpile is an analyst estimate.
- The 25% charge on H200 sales is described as a revenue share by some sources and as a Section 232 tariff by others.
- Smuggling volumes are unknown; only prosecuted cases are public.
- Calculators and scenarios are illustrative.
๐ References
- US rules and filings
- Federal Register: Revision to License Review Policy for Advanced Computing Commodities (15 Jan 2026)
- BIS: Commerce revises license review policy for semiconductors exported to China
- Nvidia fiscal 2026 Form 10-K (China licensing)
- H200 and MI325X China licences go case-by-case
- DailyTechWire: Nvidia's H200 China sales
- Huawei, SMIC and the Chinese stack
- Epoch AI: Will Huawei catch up to Nvidia by 2030?
- Huawei: Groundbreaking SuperPoD interconnect (Connect 2025 keynote)
- RCR Wireless: Huawei outlines roadmap for Ascend AI chips
- Tom's Hardware: SMIC 5nm for Huawei (2025)
- ChinaTalk: Mapping China's HBM advances
- AI Frontiers: High-bandwidth memory, the critical gaps in US export controls
- TechPowerUp: CXMT plans HBM3 line in 2026
- CXMT reportedly starts small-scale HBM3E production (Aug 2026)
- Design & Reuse: China's AI chip ambitions limited by HBM supply
- Caixin: Cambricon's breakneck growth starts to slow
- Cambricon Q3 2026 results
- CNBC: MetaX and Moore Threads IPOs
- EE Times: China EUV breakthrough
- Implicator: China's EUV machine โ what it can and can't do yet
- Smuggling and workarounds
- Bloomberg Government: Man charged with illegally shipping Nvidia chips to China (Oct 2026)
- AP: Taiwan prosecutors investigate Nvidia chip smuggling
- DeepSeek-AI: Insights into DeepSeek-V3 โ scaling challenges and reflections on hardware
- DeepSeek-AI: DeepSeek-V3 technical report
- Related on this site
- The Chokepoint ยท The Memory Wall ยท The Interconnect Wars ยท China AI ecosystem