Nvidia H200 chips are back in focus after the United States quietly approved additional Chinese companies to purchase the advanced AI processors. The decision marks another shift in Washington’s approach to AI exports, as the U.S. continues balancing national security concerns with the commercial interests of American chipmakers.
Reuters reported that, ZTE Kangxun Telecom, server manufacturer Maginfra, and Zhuhai Hengqin Yunxiang Zhisheng Network Technology—a subsidiary of cloud computing company Kingsoft—have all received authorization to access advanced AI hardware. ZTE Kangxun Telecom and Maginfra have been approved to purchase Nvidia’s H200 chips, while Kingsoft’s subsidiary has been cleared to use competing AI accelerators from AMD.
The approvals expand the list of Chinese companies allowed to acquire cutting-edge AI processors. Earlier this year, Reuters reported that major technology firms including Alibaba, Tencent, ByteDance, and JD.com had also received export licenses, although shipments had not yet moved forward because of additional reviews in both Washington and Beijing.
Which Companies Can Buy Nvidia H200 Chips?
The newly approved companies represent a broader range of China’s technology industry beyond the country’s largest internet giants.
ZTE Kangxun Telecom, a subsidiary of telecommunications equipment manufacturer ZTE, and server company Maginfra have both been granted permission to purchase Nvidia’s H200 AI chips. Meanwhile, Zhuhai Hengqin Yunxiang Zhisheng Network Technology, a subsidiary of Kingsoft, has received approval to acquire AMD AI accelerators designed to compete directly with Nvidia’s H200.
While none of the companies have publicly commented on the approvals, their inclusion suggests that the U.S. licensing process is gradually expanding beyond a handful of major cloud providers.
Why the Nvidia H200 Matters
The Nvidia H200 is one of the company’s most powerful AI accelerators and plays a critical role in training and running large language models. It offers significantly higher memory capacity and improved performance compared to previous generations, making it particularly valuable for cloud providers, AI research labs, and companies developing generative AI systems.
Demand for advanced AI chips has surged worldwide as organizations race to build larger and more capable artificial intelligence models. Nvidia has become the dominant supplier in this market, with its GPUs powering many of the world’s leading AI platforms.
Because of its capabilities, the H200 has also become one of the most closely monitored pieces of technology in the ongoing competition between the United States and China.
Why Is the U.S. Allowing Some AI Chip Sales to China?
Since 2022, Washington has steadily tightened export controls on advanced semiconductors, arguing that unrestricted access to high-performance AI hardware could strengthen China’s military capabilities and accelerate the development of sensitive technologies.
However, policymakers also face growing pressure from the U.S. semiconductor industry. Nvidia has repeatedly argued that completely cutting off China from American AI technology would not stop Chinese AI development but could instead encourage domestic competitors to become stronger while reducing revenue for U.S. companies.
The latest approvals suggest the Trump administration is trying to balance those competing priorities. Instead of reopening exports broadly, regulators appear to be allowing carefully controlled sales through individual export licenses while maintaining restrictions on the most advanced technologies.
This approach allows the U.S. government to retain oversight of who can purchase AI hardware while preserving access to one of the world’s largest technology markets.
Shipments Still Face Uncertainty
Receiving an export license does not automatically guarantee that the chips will reach Chinese customers.
According to Reuters, some Chinese cloud providers have recently informed partners that H200 chips may soon become available, suggesting that import approvals inside China are also progressing. However, Chinese authorities have increasingly encouraged domestic companies to adopt locally developed AI hardware instead of relying on foreign suppliers.
As a result, approved sales could still face delays or additional regulatory reviews before deliveries are completed.
The uncertainty reflects the increasingly complex nature of the global AI supply chain, where political decisions now play nearly as important a role as technological innovation.
What Nvidia H200 Chips Mean for the Global AI Race
For Nvidia, maintaining access to the Chinese market remains strategically important. China represents one of the world’s largest buyers of AI computing infrastructure, and continued sales could generate billions of dollars in revenue while helping the company maintain its leadership position.
For the broader AI industry, the latest approvals highlight a changing approach to export controls. Rather than relying on blanket restrictions, governments are increasingly choosing selective licensing, allowing certain companies to purchase advanced hardware while continuing to limit access for others.
As competition between the United States and China continues to intensify, Nvidia H200 chips are becoming more than just powerful AI hardware. They have become a strategic technology that could shape the future of artificial intelligence.
Source: Reuters
