Nvidia is developing a version of its Groq-based AI inference chips specifically for the Chinese market, with availability expected as early as May 2026, according to Reuters sources. This strategic move follows the company's recent acquisition of U.S. export licenses from the Trump administration, allowing it to resume production and sales of its H200 processors to Chinese customers.
The development comes alongside Nvidia CEO Jensen Huang's significant endorsement of OpenClaw technology at the GTC 2026 developer conference in San Jose, California. Huang publicly described OpenClaw as "definitely the next ChatGPT" and framed it as a foundational computing layer comparable to Windows, Linux, Kubernetes, and HTML. His comments triggered immediate market reactions, with Chinese AI stocks like MiniMax and Zhipu surging more than 20% on Wednesday as investors interpreted his statements as a green light for OpenClaw-related investments.
The Groq chips bound for China are not downgraded versions and are not manufactured specifically for the Chinese market, according to sources. However, they can be adapted to work with other systems. This contrasts with Nvidia's upcoming Vera Rubin processors, which face U.S. export controls that prohibit sales to China. For domestic and non-restricted markets, Nvidia plans to pair Groq chips with Vera Rubin processors.
Nvidia's push into the Chinese inference chip market represents a competitive response to local manufacturers. Major Chinese companies, including Baidu, already produce their own inference chips. The Groq technology, which Nvidia licensed from AI chip startup Groq in a $17 billion deal late last year, targets the AI inference segment—where trained models generate responses, write code, or perform tasks in real time.
Simultaneously, Chinese authorities are actively building an ecosystem around OpenClaw technology. Local reports from Shenzhen's Longgang district reveal officials discussing "Lobster Service Zones" for OpenClaw and offering subsidies covering up to 50% of data-governance costs and approximately 30% for specific AI storage devices.
While Huang's endorsement has provided a fresh boost to China's AI trade, market observers note that sustained momentum will depend on whether companies can translate hype into genuine usage and solid revenue. MiniMax, for instance, generated just $79 million in revenue in 2025 despite its market value surpassing HK$380 billion following a 28% single-day jump on March 10.
