Nvidia shares climbed to a new all-time high Thursday after reports said the US cleared around 10 Chinese companies to buy its H200 AI chips, reopening a major sales opportunity in one of the world’s most important AI markets.
The stock rose nearly 4% Thursday and was trading near $234 at press time, helping lift Nvidia’s market value to about $5.7 trillion as the broader tech rally pushed the S&P 500 and Nasdaq to record intraday highs. The move extended Nvidia’s record run after the company crossed the $5.5 trillion market cap threshold earlier this week.
The US Commerce Department has approved around 10 Chinese companies, including Alibaba, Tencent, ByteDance, and JD.com, to purchase Nvidia’s H200 chips. Distributors including Lenovo and Foxconn have also received approvals, and each approved customer can buy up to 75,000 chips under the licensing terms.
No H200 deliveries have been made so far, leaving the deal in limbo despite the approvals. The delay reflects continued tension between Washington’s export control regime and Beijing’s push to reduce reliance on US technology while building domestic AI chip capacity.
The development comes as Nvidia CEO Jensen Huang joins a US delegation in Beijing during high-level talks between President Donald Trump and Chinese President Xi Jinping. Huang was added to the trip after an invitation from Trump, raising investor hopes that the visit could help unlock the stalled chip sales.
The H200 is Nvidia’s second most powerful AI chip and remains a key product for training and running advanced AI systems. The approvals could restore part of Nvidia’s access to China, a market where the company’s sales have been constrained by export restrictions aimed at limiting Beijing’s access to advanced AI hardware.
Nvidia’s rally also reflects Wall Street’s continued confidence in AI infrastructure spending ahead of the company’s upcoming earnings report. Analysts have raised price targets in recent days, citing the expansion of the AI data center market and Nvidia’s growing role as a core supplier of compute capacity.