Nvidia to ship AI chip for China by year-end, The Information reports

Nvidia has dismissed a report by The Information suggesting it will launch an AI chip specifically for Chinese customers by the end of the year. According to Nvidia, there is no China-specific version of its Language Processing Unit (LPU) in its current roadmap, and the company has not shipped any LPU products to customers in China.

The Reuters news agency had previously reported in March that Nvidia was developing a China-compatible version of its AI chips to compete in the rapidly expanding AI inference market. The latest report by The Information, citing two unnamed Nvidia employees, claimed that the company would begin small-scale shipments of the China-focused AI chip by the end of the year.

The chip, based on Nvidia's LPU technology licensed from Groq, was said to work alongside graphics processing units (GPUs) to enhance the speed of AI chatbot responses. The publication also noted that the chip complied with U.S. export control regulations and that Nvidia had made software modifications to enable the LPU to function with processors available in China following the unavailability of Nvidia's next-generation Vera Rubin AI system due to U.S. restrictions.