Nvidia H200 China deal survives Trump-Xi summit, but not as expected

President Trump traveled to Beijing with Jensen Huang joining the delegation at the eleventh hour, departing two days later with vague remarks about potential chip export developments. No deal materialized. Since Trump first authorized sales in December 2025, not a single Nvidia H200 has reached China. US Trade Representative Jamieson Greer confirmed to Bloomberg that semiconductor restrictions were not even discussed on the bilateral agenda.

Beneath the summit’s pageantry lay a more significant shift. The H200 is not stalled due to Washington’s refusal; the US has already granted permission. Approximately 10 major Chinese firms, including Alibaba, Tencent, ByteDance, and JD.com, hold approved US export licenses for up to 75,000 units each, with Lenovo and Foxconn authorized as distributors. The chips remain undelivered because Beijing prohibits its companies from accepting them.

Two frameworks, one deadlock

Understanding the mechanics of this stalemate is crucial. US regulations mandate that all H200 chips ordered by Chinese clients must be used exclusively within China. Conversely, Beijing has instructed domestic tech firms to restrict Nvidia chip usage to overseas operations while bolstering local manufacturing. These two requirements are fundamentally incompatible.

Chips cleared for export cannot legally be deployed in the locations Beijing desires, and Beijing refuses to authorize the domestic usage required by US licenses, according toImplicator.

Commerce Secretary Howard Lutnick noted at a Senate hearing last month that Chinese firms are prioritizing investments in domestic suppliers like Huawei. The State Council has also ordered a supply-chain security review designed to reduce reliance on US semiconductors.

This policy contradiction is intentional. That is the objective.

What Huawei gained while diplomats talked

Events surrounding the summit yielded data points more significant for the long term than Trump’s final remarks. DeepSeek confirmed its latest model was optimized for Huawei processors. Tencent’s chief strategy officer stated that Chinese GPU supply will expand progressively through 2026, while an Alibaba executive reported that its T-Head proprietary GPUs have achieved mass production at scale.

This follows the April launch of DeepSeek V4, which adapted the model for Huawei’s Ascend chips—the first major Chinese frontier model to do so during training, not just inference. The summit week confirmed that this transition is no longer experimental; it is now official supply-chain policy. Nvidia’s China revenue has dropped to roughly 5% in recent quarters, down from over 20% before export controls tightened. The company’s guidance for the current quarter assumes zero revenue from China.

Huang’s last-minute addition to the delegation—Trump called him directly after seeing media coverage that he had not been invited—signaled urgency. The outcome highlighted the limits of CEO diplomacy when the obstruction is structural rather than procedural.

The read for the AI industry

This stalemate extends beyond bilateral optics. Chinese AI platforms are now operating under a domestic mandate to build on Huawei’s compute stack. The question of which AI hardware architecture will dominate the world’s second-largest AI market is being resolved not by technical benchmarks but by government directive.

Beijing’s push toward Huawei Ascend chips instead of Nvidia H200S is more than a trade stance. It is a structural bet that the performance gap will close quickly enough to make being locked into the domestic stack manageable. DeepSeek V4’s results suggest this may be accurate, at least for inference workloads.

Trump hinted that something could happen. Greer stated the decision is sovereign for China. Both statements are true, yet neither alters the current reality: the H200 deal is approved, licensed, and frozen, with Huawei filling the void it leaves behind.

(Image source: The White House)

See Also: Can China’s chip stacking strategy really challenge Nvidia’s AI dominance?

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.