NVIDIA has received U.S. government approval to ship a limited number of its H200 AI chips to customers in China, but the company has not booked a single dollar of revenue from those shipments and has kept any China-related data centre income completely out of its first-quarter guidance. The situation reflects deep uncertainty on both sides of the Pacific — Washington has approved the exports with conditions, but Beijing has not yet confirmed it will allow the chips to enter the country. As a result, one of the most significant potential revenue streams in the global AI chip market remains in limbo.
What Washington Approved — and What It Comes With
The U.S. government issued Nvidia a license allowing it to ship H200 chips to Chinese customers, but the approval comes with notable restrictions. According to Bloomberg News, the chips are subject to physical inspection in the United States before they can leave, and they incur a 25% import duty on top of that. These conditions significantly reduce the commercial attractiveness of the arrangement compared to unrestricted exports.
The H200 is Nvidia’s high-performance AI accelerator based on the Hopper architecture. It sits below the company’s flagship Blackwell chips in capability but remains a powerful tool for AI training and inference workloads. China had been a significant market for Nvidia before a series of escalating U.S. export controls, starting in 2022, progressively cut off access to the company’s most advanced chips.
Beijing’s Response Is the Missing Variable
Even with a U.S. license in hand, Nvidia faces a second and equally uncertain gating factor: China itself. The Chinese government has not yet indicated whether it will permit imports of the H200 chips under the new conditions, and Nvidia’s CFO Colette Kress made clear on Wednesday’s earnings call that the company simply does not know how Beijing will respond.
Because of that uncertainty, Nvidia deliberately excluded any China data centre revenue from its first-quarter fiscal 2027 outlook of approximately $78 billion. That figure, already a record forward guidance number, was built entirely without factoring in any potential H200 sales to China. If Beijing does ultimately open the door, any resulting revenue would represent genuine upside to current estimates.
A Warning About Chinese Competitors
Beyond the immediate revenue question, Kress used the earnings call to deliver a broader strategic warning about the competitive landscape in China. She pointed out that Chinese AI chip companies, recently boosted by a wave of domestic IPOs, are gaining ground and have the potential to reshape the global AI compute market over the long run.
Her message was pointed: if the U.S. wants to maintain its dominance in AI infrastructure, it needs to be able to compete everywhere — including in China. Restricting American companies from serving Chinese customers, she argued, risks ceding that ground to homegrown Chinese alternatives that are advancing quickly.
This is a tension that U.S. policymakers have struggled to resolve. Export controls are designed to keep cutting-edge AI technology out of Chinese military hands, but overly broad restrictions also push Chinese companies and developers toward building and using domestic alternatives — potentially accelerating exactly the competitive threat the controls are meant to prevent.
Market Implications
For investors, the H200 China situation is best understood as a potential call option embedded in Nvidia’s already-strong outlook. The base case — no China revenue — is already priced into Q1 guidance. Any positive resolution, whether Beijing approves imports or U.S. policy evolves to allow easier access, would add incremental upside that is not currently reflected in consensus estimates.
At the same time, the growing strength of Chinese domestic chip companies is a longer-term risk worth monitoring. Names like Huawei and a new crop of well-funded AI chip startups are actively developing alternatives to Nvidia’s hardware, with government backing and a large domestic market to prove out their technology.
What to Watch
Investors should keep a close eye on three developments. First, watch for any official response from Beijing regarding whether H200 imports will be permitted — a green light from China would immediately add a new revenue layer to Nvidia’s already-record guidance. Second, monitor the trajectory of Chinese AI chip companies that recently completed IPOs; if they begin winning enterprise customers at scale, that could begin to erode Nvidia’s addressable market in Asia. Third, watch for any changes to U.S. export control policy under the current administration — the chip licensing landscape has shifted multiple times in recent years, and further adjustments in either direction could materially change Nvidia’s China revenue outlook. Any formal trade negotiations between Washington and Beijing would also be a key catalyst to monitor.
