(Bloomberg) -- Z.AI has completed construction of a major data center that it plans to fill only with Chinese-made chips, a step forward in Beijing's efforts to shift away from restricted Nvidia Corp. silicon for future AI development.
Most Read from Bloomberg
The company formerly known as Zhipu has begun partially operating a 1-gigawatt hub designed to help the Chinese company develop and deploy its cutting-edge GLM platforms, a person familiar with the matter said. At full utilization, the facility would consume enough power to energize roughly 750,000 homes at any given moment.
Z.AI has installed at least 10,000 Chinese-made chips there, the person said, without specifying how much overall capacity has been deployed. Z.AI has now built or operates several computing clusters each with more than 10,000 locally designed semiconductors, located at this data center and others, the person said, asking to remain anonymous to discuss non-public information.
Z.AI shares closed up 37% on Tuesday. Representatives for the company didn't respond to an emailed request for comment.
At a 1-gigawatt scale, Z.AI's data center would be among the largest server hubs built by a Chinese AI lab. Cloud providers such as Alibaba Group Holding Ltd. and China Telecom Corp. are known to be the biggest builders of computing infrastructure currently, given the billions of dollars required not just to purchase high-end servers in bulk but also to ensure stable power and water supply.
If Z.AI were to completely fill a 1-gigawatt facility with domestic processors, it would need to procure hundreds of thousands of chips from companies like Huawei Technologies Co. or Cambricon Technologies Corp., China's top competitors to Nvidia. Bloomberg News reported last year that Huawei plans to make around 1.6 million Ascend dies in 2026: enough base components for around 600,000 of the company's marquee 910C chips plus a smattering of other models. Cambricon, meanwhile, was targeting half a million AI accelerators this year. Questions have persisted around how Chinese AI labs like Moonshot and DeepSeek developed some of the world's most capable AI models, and whether local alternatives to Nvidia are capable enough to support longer-term development of the technology.