The Trump-Xi summit in Beijing ended without the breakthrough investors had hoped for on Nvidia’s H200 chip sales to China, leaving a potentially large revenue stream unresolved. Despite the Trump administration’s January approval for conditional H200 exports, the two-day meeting produced no progress on the practical barriers that still block shipments. U.S. Trade Representative Jamieson Greer said semiconductor export controls never became a major focus of the talks, a clear sign that bigger strategic issues took precedence.

That outcome matters because the stakes are unusually high. Chinese companies including Alibaba, Tencent, ByteDance and JD.com have already received U.S. approval to buy the chips, but Beijing has yet to grant the domestic import clearances needed to complete those orders. President Trump said after the summit that China was holding back because it wants to build its own alternatives, underscoring how commercial demand is colliding with industrial policy.

The numbers help explain the market’s sensitivity. Chinese technology groups have reportedly placed orders for more than 2 million H200 chips at roughly $27,000 each, implying about $54 billion in potential sales, well above Nvidia’s current available supply. Nvidia shares had climbed to a record before the summit on expectations that the licensing framework might finally translate into shipments. The lack of movement now puts added pressure on management to address the China question when it reports earnings on May 20.

The uncertainty is significant, but not entirely new. Nvidia said in its latest quarterly report that it had permission to ship only small amounts of H200 products to selected Chinese customers and had not yet booked any revenue under the program. The company also said it still did not know whether imports would ultimately be allowed into China. That caution suggests management has been careful not to build a reopening of the Chinese market into near-term expectations.

Even so, the financial upside remains material. Analysts estimate broader access to China could add as much as $26 billion in annual revenue for Nvidia at a time when global AI infrastructure spending continues to expand. The company posted quarterly revenue of $46.7 billion in its most recent report, with data center sales up 56% from a year earlier, showing that demand elsewhere remains strong.

The bigger issue is what this impasse says about the direction of U.S.-China technology ties. Washington’s export framework was never a full reset; it came with strict conditions, including revenue-sharing with the U.S. government and continuing oversight of shipments. Beijing, for its part, appears increasingly committed to reducing dependence on American chips by backing domestic suppliers such as Huawei and SMIC. That makes the H200 dispute less a licensing delay than a test of whether either side still sees room for selective cooperation in critical technologies.

For investors, the immediate risk is straightforward: hopes for a quick China-related boost to Nvidia’s revenue may need to be pushed out again. Still, the company’s broader growth story remains intact, supported by strong demand for AI servers, networking products and data center infrastructure outside China. Unless lower-level negotiations gain traction soon, the H200 program is likely to remain stuck in administrative and political limbo, with the next signal coming not from a summit stage but from regulatory follow-through on both sides.