Jensen Huang says “we don't know” if China would accept Nvidia's H200 AI chips should the US relax export controls on them, after a meeting with President Trump
FT Report Flags Pivotal AI Chip Export Decision for Traders The Korea Herald : Trump meets Nvidia CEO Huang: source DigiTimes : Taiwan's president invokes past chip curbs as US weighs Nvidia sales to China Gideon Wolf / COINOTAG : Nvidia CEO Huang: China Likely to Reject H200 Chips Even if US Eases Export Rules Naomi Li Gan / Tech in Asia : Nvidia CEO unsure China would take its H200 chips Deborah Sophia / Reuters : Trump met with Nvidia CEO Jensen Huang about export controls, CBS News' reporter says Yifan Yu / Nikkei Asia : Nvidia's Jensen Huang warns of China's AI ‘belt and road’ ambitions Alana Mastrangelo / Breitbart : Nvidia CEO Jensen Huang Praises Trump on Joe Rogan Podcast: ‘Very Practical, Common Sense, and Logical’ Anton Shilov / Tom's Hardware : Nvidia lobbies White House and wins loosened AI GPU export control to China — U.S. lawmakers reportedly reject GAIN AI Act South China Morning Post : Nvidia's Huang not sure if China would accept its H200 chips, he says after Trump meeting
Context & Ripple Effects
The meeting followed a period in which Nvidia had already secured a return of H20 sales through Huang’s lobbying of the administration, while officials were considering a far larger policy shift: the resumption of H20 sales and discussions over H200 access.
The key constraint is not solely Washington’s licensing posture. Related coverage later indicated Beijing was preparing to restrict H200 access, and subsequent reporting found no H200 shipments had occurred, underscoring a gap between export authorization and actual sales.
First-order effects
- Nvidia cannot treat a potential US relaxation as a guaranteed China revenue opportunity; customer acceptance becomes a separate commercial and policy risk.
- US officials weighing H200 rules face a less straightforward trade-off: authorizing exports may not translate into deliveries if Chinese access is constrained.
Second-order effects
- Chinese AI infrastructure buyers gain another reason to plan around non-Nvidia supply paths, rather than wait for a reversible US policy decision.
- Nvidia’s China product strategy faces a two-sided compliance problem: chips must satisfy US export rules while remaining acceptable under Chinese procurement and self-sufficiency priorities.
Third-order effects
- If this pattern persists, advanced-chip trade will be governed by overlapping US export controls and Chinese market-access decisions, reducing the ability of either side’s policy change to determine outcomes alone.
- The episode points toward AI infrastructure sovereignty: governments increasingly shape not just where compute can be sold, but which hardware domestic users can depend on over time.
The trend: AI-chip competition is shifting from a US licensing question into a two-sided sovereignty contest over dependable access to advanced compute.