Nvidia has stopped producing H200 AI chips meant for China, reallocating capacity to next-generation processors amid tightening US export controls and Beijing’s push for domestic semiconductors
US semiconductor giant Nvidia has stopped producing its H200 artificial intelligence (AI) chips intended for the Chinese market, the Financial Times reported on Thursday, signalling a fresh escalation in the technology tussle between United States and China.
According to the report, the company has shifted manufacturing capacity at Taiwanese chip foundry Taiwan Semiconductor Manufacturing Company (TSMC) away from H200 processors to its next-generation Vera Rubin architecture.
The move suggests Nvidia does not expect meaningful sales of the H200 chips in China in the near term as regulatory barriers on both sides complicate shipments of advanced semiconductors.
Export curbs cloud China sales
The H200 processor is among Nvidia’s older-generation AI chips and was designed to comply with US export restrictions on advanced semiconductor technology. However, shipments have remained uncertain as Washington continues to tighten rules governing the export of high-performance chips to China.
A US Commerce Department official said last month that none of Nvidia’s H200 chips had been sold to Chinese customers so far.
Earlier this year, the administration of Donald Trump formally allowed limited sales of the chips to China. But shipments have remained stalled due to regulatory guardrails and licensing requirements.
Last week, Nvidia said it had received licences from the US government allowing it to ship only “small amounts” of H200 processors to customers in China.
Manufacturing shifts to next-generation chips
Industry sources cited in the report said Nvidia had redirected production capacity to its new Vera Rubin chip platform, unveiled earlier this year to power more advanced AI systems.
Demand for the new architecture has been strong from leading US technology companies including OpenAI and Google.
Analysts say the shift reflects Nvidia’s strategy to prioritise markets where demand and regulatory clarity are stronger.
“Instead of waiting in limbo, Nvidia has to move on to what it can achieve with certainty, especially when there is a shortage of supply for its most advanced products,” a person familiar with the plans told the Financial Times.
China pushes domestic semiconductor drive
At the same time, Beijing is stepping up efforts to reduce reliance on foreign chipmakers and boost domestic semiconductor capabilities.
China is considering restrictions on purchases of H200 processors in a bid to support local chipmakers and encourage AI companies to adopt homegrown technology.
Customs authorities have reportedly listed H200 chips among restricted imports unless companies obtain special approval from Beijing.
Despite US curbs, analysts say China continues to make steady progress in artificial intelligence and semiconductor development as government-backed firms accelerate investment in domestic chip design and manufacturing.
Geopolitics reshaping the AI chip market
The production halt underscores how geopolitical tensions are reshaping the global semiconductor industry, particularly in the rapidly expanding AI chip market.
Nvidia had previously expected Chinese customers to place orders for more than one million H200 units, with suppliers preparing for shipments earlier this year.
However, prolonged uncertainty over export approvals and possible Chinese countermeasures appears to have forced the company to pivot its manufacturing plans.
Some analysts believe the issue could still feature in diplomatic discussions between Beijing and Washington, with speculation that leaders including Xi Jinping and Trump could address technology trade restrictions in future talks.
For now, the shifting production strategy highlights the growing impact of geopolitical rivalries on the global race to dominate artificial intelligence infrastructure.
End of Article
Leave a Reply