Samuel Boivin | Nurphoto | Getty Images
Alibaba and China Telecom are launching a data center in southern China powered by the e-commerce giant’s own chips, as the country ramps up its focus on homegrown AI infrastructure.
The facility, announced on Tuesday, will feature 10,000 of Alibaba’s Zhenwu semiconductors which are designed for AI training and inferencing along with the ability to support AI models the size of hundreds of billions of parameters.
These are among some of the largest models out there and underscore how China’s biggest tech players are advancing their own AI semiconductor technology as Beijing intensifies its push for self-sufficiency.
Over the past few years, the U.S. has looked to restrict China’s access to key semiconductor technology, including AI chips from Nvidia, which has accelerated the country’s efforts to develop domestic alternatives.
Alibaba, one of China’s largest tech firms, has been designing its own chips through its T-head unit. The Hangzhou-headquartered company is also one of China’s biggest cloud computing players. It designs chips, builds data centers, and develops its own AI models which it then sells through its cloud computing division. Cloud computing has been among its fastest-growing businesses in recent quarters.
There is an increased focus on building large-scale data centers in China with domestic technology. Last month, a computing cluster built with Huawei’s advanced Ascend 910C AI chips went online.
While U.S. tech giants are expected to spend around $700 billion this year to fuel their AI build-outs, Chinese companies have taken a different approach. They are spending less and have focused their AI on industries they believe will drive revenue growth and return on their investments.
China Telecom and Alibaba said the data center, located in Shaoguan, in the Guangdong province of China, is expected to expand to a scale of 100,000 chips. The computing cluster can be used in industries from healthcare to advanced materials, they added.
