Trump Approves Nvidia’s H200 Exports to China in Strategic Shift as Analysts See Mixed Motives
US President Donald Trump’s decision to allow Nvidia to export its H200 artificial intelligence chips to approved customers in China marks a notable shift in Washington’s technology policy, opening limited market access while still restricting cutting edge hardware. Analysts say the move represents a calculated middle path, one intended to preserve US commercial influence while tempering China’s push for technological self sufficiency.
The change, announced by Trump on Monday, permits Nvidia to sell the H200 under conditions that he said would ensure strong national security oversight. He also said that twenty five percent of the value would go to the US, though details on how the fee structure will work have not yet been released. The approval reportedly extends to other American chipmakers, including AMD and Intel, though it explicitly excludes Nvidia’s newest and most advanced Blackwell and Rubin series chips.
Experts describe the policy as a sliding scale approach designed to balance competing US priorities. On one hand, Washington aims to curb China’s access to top tier AI training chips that could advance military or strategic capabilities. On the other, US companies have warned that total exclusion from the Chinese market risks long term commercial losses that could undermine America’s dominance in the semiconductor sector. The H200, while powerful, is no longer Nvidia’s most sensitive technology, allowing it to be positioned as a compromise tool.
Analysts say the decision may help retain US suppliers’ foothold in China at a time when Chinese firms are accelerating development of domestic alternatives. Beijing has expanded investment in homegrown semiconductor technologies, backed by companies like Huawei, Cambricon, Moore Threads and MetaX. Allowing the sale of slightly older AI chips could slow the urgency behind China’s efforts while still blocking access to the highest performance products.
However, the green light is already stirring debate within the US policy community. Some lawmakers and national security officials argue that loosening restrictions, even selectively, risks giving China technology that could be repurposed for sensitive uses. Others contend that overly harsh export bans ultimately accelerate China’s drive for self reliance and hurt American firms in the process.
The tug of war is expected to continue as Washington weighs how to manage competition with China while protecting US economic interests. Nvidia, which has relied heavily on China as a major customer base, has been pushing for clarity and flexibility in export rules. The company has previously redesigned chips to comply with US regulations, but the broader uncertainty has complicated long term business planning.
For China’s AI sector, access to the H200 offers some short term relief. Many developers have struggled with shortages following earlier restrictions, and the H200 remains competitive in performance despite being surpassed by Nvidia’s latest generation chips. Analysts expect Chinese companies to move quickly to secure approved shipments while continuing to invest heavily in local chip development to avoid future vulnerability.
The policy marks a rare moment of partial reopening in an otherwise tense US China tech landscape. How long the window remains open will depend on political dynamics in Washington, industry lobbying and the pace of China’s technological progress.
