Nvidia’s Zero China Share: The $8 Billion Admission That Export Controls Backfired
Jensen Huang's stark disclosure exposes both the collapse of US semiconductor dominance and the failure of containment policy to slow Chinese chip advancement.
Nvidia CEO Jensen Huang confirmed in October 2025 that his company’s market share in China’s advanced AI accelerator market collapsed from 95% to zero, erasing what had been 20-25% of the firm’s data center revenue and marking the most significant corporate casualty of US export control policy. The admission, disclosed at a Citadel Securities event, represents an $8 billion quarterly revenue loss and crystallises a strategic paradox: controls designed to kneecap Chinese AI capability instead accelerated domestic semiconductor development while handing Beijing proof that technology containment works against American interests.
The financial damage is concrete. Nvidia recognised $4.5 billion in H20-related charges in Q1 fiscal 2026 (ended January 2026). China/Hong Kong revenue plunged 63% to $3 billion in Q3 fiscal 2026, per the South China Morning Post, with H20 sales — Nvidia’s export-compliant chip — generating only $50 million. Huang’s framing was blunt: “I can’t imagine any policymaker thinking that that’s a good idea, that whatever policy we implemented caused America to lose one of the largest markets in the world.”
The Strategic Backfire
Export Controls that began in 2022 under the Biden administration and intensified through 2025 rested on a premise Huang now openly challenges: that denying China access to cutting-edge chips would prevent AI advancement. Instead, data from CSIS shows domestic chipmakers are forecast to capture 50% of China’s AI chip market by 2026, up from marginal single-digit shares in 2023. Huawei’s AI chip market share alone now exceeds 50% in China, filling the void Nvidia left.
“The U.S. has based its policy on the assumption that China cannot make AI Chips. That assumption was always questionable and now it’s clearly wrong. China has enormous manufacturing capability.”
— Jensen Huang, CEO, Nvidia
The manufacturing evidence supports Huang’s assessment. American Enterprise Institute research published in March 2026 confirms China’s existing DUV lithography fleet can sustain several million 7nm wafer starts annually — enabling advanced chip production without EUV equipment that export controls were designed to deny. SMIC, China’s largest foundry, reported record 2025 revenue of $9.3 billion (up 16%) and is forecast to reach $11 billion in 2026, driven by domestic AI chip orders.
The Self-Sufficiency Acceleration
Chinese GPU startups collectively known as the “four little dragons” — Moore Threads, Biren Technology, Iluvatar CoreX, and MetaX — have evolved from marginal players to credible alternatives. Moore Threads guided 2025 revenue of 1.45-1.52 billion yuan, a 231-247% year-over-year increase, according to CNBC. Industry estimates suggest China could meet 80% of AI semiconductor demand with local supply, a threshold that would render future export restrictions largely irrelevant.
Government backing explains the velocity. Beijing’s five-year semiconductor plan, unveiled in 2024, directed subsidies toward AI chip development with explicit targets for import substitution. The result is not merely incremental improvement but a compression of the development timeline that would have taken a decade under market conditions alone. According to industry analysis, Chinese AI chips have evolved from “usable” to “easy to use.”
Geographic Concentration Risk
Nvidia’s China revenue collapse amplifies concentration in its remaining markets. With China eliminated, the company depends almost entirely on US hyperscalers (Amazon, Microsoft, Google, Meta) and select Asian buyers for data center growth. This creates sovereign risk exposure: any future regulatory action in Europe or allied economies could replicate the China shock. Huang’s public criticism of export policy — rare for a CEO of a firm deeply embedded in US defense supply chains — signals awareness that geographic diversification is now a strategic imperative Nvidia cannot achieve.
- Export controls eliminated Nvidia’s China revenue without slowing Chinese AI chip development, creating the worst outcome for US policy objectives.
- Chinese domestic chipmakers now hold 50%+ market share in their home market, up from negligible levels in 2023, with 80% self-sufficiency forecast by 2027.
- US semiconductor firms face permanent geographic concentration risk as China demonstrates technology containment accelerates rather than prevents indigenous capability development.
- Huang’s testimony provides corporate validation for policy reassessment across administrations, framing controls as counterproductive to American competitiveness.
The policy debate now fractures along familiar lines. Council on Foreign Relations analysis argues controls remain justified despite Chinese advancement, noting Huawei’s chips still lag Nvidia’s on performance-per-watt and ecosystem maturity. The counterargument, which Huang articulates, is that conceding “an entire market the size of China probably does not make a lot of strategic sense” when the primary goal — preventing Chinese AI capability — demonstrably failed.
What to Watch
Nvidia’s Q2 fiscal 2027 earnings (expected late summer 2026) will quantify whether the China revenue loss stabilises or continues to degrade as export-compliant chips face domestic alternatives. Congressional testimony from Huang, likely in coming months given the policy implications, could force Commerce Department officials to defend export controls against the CEO’s evidence that they backfired. Watch for European regulatory moves — any EU export restrictions mirroring US policy would validate Nvidia’s concentration risk thesis and potentially trigger a broader reassessment of technology containment as viable strategy. Chinese chipmaker capacity announcements through year-end will confirm whether the 80% self-sufficiency estimate proves conservative or optimistic, determining whether US export policy created a temporary disruption or permanent strategic defeat.