Chief Executive Jensen Huang had said in May that Nvidia has largely conceded China’s vast AI-chip market to Huawei Technologies, which has emerged as the company’s main domestic rival.
Nvidia, on Thursday denied a report by The Information that it would be shipping a Language processing unit (LPU), tailored for Chinese customers by the year-end, saying that it has no China-specific roadmap.
Earlier, on Thursday, The Information reported, citing two Nvidia employees, that the company plans a small set of shipments of artificial intelligence (AI) chips designed for Chinese customers by the year-end and that several customers had already placed orders.
“The reporting in The Information on Nvidia’s LPU is incorrect. We have no LPU sales in the China market today, and no China-specific LPU product in our roadmap,” an Nvidia spokesperson said.
Reuters had also reported in March that Nvidia is working on a version of AI-chips compatible with Chinese regulations to get a foothold in the vast AI inference market in the country.
According to The Information’s report, that Nvidia’s LPU is designed using technology, licensed from startup Groq that works alongside Graphic Processor Units (GPUs) to speed up AI chatbot responses.
The publication said that the new chip complied with the US export restrictions and the LPU will work with processors available in China, after Nvidia’s next-gen Vera rubin AI was banned by the US government for export to China.
Nvidia’s business in China is passing through a phase of uncertainty.Washington in May gave approval to the company to sell its H200 AI chips to select Chinese customers such as Tencent, Alibaba, and ByteDance. Meanwhile the company is also marketing its Vera processors to Chinese customers.
Chief Executive Jensen Huang had said in May that Nvidia has largely conceded China’s vast AI-chip market to Huawei Technologies, which has emerged as the company’s main domestic rival.
