consultantsreview logo

Consultants Review Magazine

It's Possible that Nvidia is Creating Another "Specially Tailored" Product for the Chinese Market

By Consultants Review Team Friday, 26 July 2024

Nvidia is creating a specially designed Al chip for China to get around US export regulations. According to a recent story, the US-based chipmaker is allegedly developing a unique server in addition to trying to stay in business in China while abiding by strict US export regulations.

According to The Information, "Nvidia has paired the chips with a special server as a new workaround to US export rules that block the sale of the advanced AI chips to China." Notably, Nvidia is introducing a new server to the market for the first time.

The article mentions two individuals who worked on the server's development and reports that Nvidia would release a new Al processor specifically for the Chinese market next year, coupled with a new server meant to optimize the chip's performance.

Nvidia could be creating AI processors tailored to China

The information was released a few days after news agency Reuters reported that Nvidia is developing an AI chip especially for the Chinese market in an effort to maintain a steady flow of income from one of its most significant markets.

According to the article, Nvidia will collaborate on the chip's introduction and distribution with Inspur, one of the American chipmaker's principal distributor partners in China.

Nvidia debuted its "Blackwell" chip series in March, along with the B200 model and the chip that is rumored to be intended for the Chinese market under the moniker "B20." Later this year, the "Blackwell" chip family will enter commercial manufacturing.

The B200 variant of the "Blackwell" chip family is said to be 30 times faster than its predecessor at various tasks, such as answering chatbot queries.

Additionally, it is stated that these chips would allow organizations to develop and execute real-time generative AI on trillion-parameter huge language models for up to 25 times less money and energy than they did with its predecessor.

Current Issue