China's Z.AI Completes 1-Gigawatt Data Center With Domestic Chips

Z.AI, the Chinese artificial intelligence company formerly known as Zhipu, has completed construction of a 1-gigawatt data center powered entirely by Chinese-made chips. The facility has started partial operations and is designed to provide the computing capacity needed to develop Z.AI's most advanced GLM model systems. The data center's power capacity is roughly equivalent to the electricity required to supply 750,000 homes at any given moment. Z.AI has also built or operated several computing clusters containing more than 10,000 chips each, indicating the company is rapidly expanding its computing infrastructure as competition among Chinese AI developers intensifies.

The project represents a significant milestone in China's broader effort to reduce reliance on restricted silicon from NVIDIA, the U.S. semiconductor company whose chips dominate AI computing. Investors may view the facility as a major test of whether China's domestic chip industry can support increasingly advanced AI models over the longer term. Huawei, China's leading designer of AI accelerators, is competing directly with Cambricon Technologies and Alibaba Group—which is both a major technology company and a cloud-computing provider—as local suppliers work to narrow the performance gap with NVIDIA. The scale of Z.AI's new facility places it among the largest data centers developed by a Chinese AI laboratory, though Alibaba and China Telecom remain among the country's largest builders of computing infrastructure overall.

Z.AI is investing in computing capacity as it faces intensifying competition from Moonshot, a Beijing-based AI startup whose Kimi K3 model has challenged leading systems from OpenAI and Anthropic. Moonshot recently suspended new subscriptions on Sunday to prioritize computing resources for existing members, illustrating the infrastructure constraints facing some Chinese AI developers. China is preparing to spend around $295 billion（約47兆円） over the next five years on data centers, and Z.AI has raised billions through a Hong Kong initial public offering and subsequent share sale. The company is on track to generate $1 billion（約1600億円） in annual recurring revenue after reaching its 2026 sales target in July, and the new data center is expected to support Z.AI's effort to strengthen its position as a provider of AI services to businesses and enterprises.

Z.AI's completion of a 1-gigawatt data center powered entirely by Chinese-made chips marks a critical inflection point in China's domestic semiconductor ambitions. The facility's power capacity—equivalent to supplying 750,000 homes—underscores the enormous computational demands of advanced AI model development. The project arrives amid intensifying competition between Chinese AI developers: Moonshot, a Beijing-based startup whose Kimi K3 model has challenged systems from OpenAI and Anthropic, recently suspended new subscriptions to prioritize computing resources for existing members, signaling acute infrastructure constraints in the sector. Z.AI's aggressive expansion—including multiple computing clusters containing more than 10,000 chips each—reflects a strategic bet that domestic semiconductor capacity can match or approximate NVIDIA's restricted offerings.

The broader context is China's commitment to reducing reliance on U.S. silicon. Huawei, Cambricon, and Alibaba are competing as local suppliers to close the performance gap with NVIDIA, and China is preparing to spend around $295 billion（約47兆円） over the next five years on data centers. Z.AI's facility may serve as a bellwether: if it successfully supports the company's path to $1 billion（約1600億円） in annual recurring revenue—a target the company aims to reach by 2026—it could validate the viability of China's domestic chip strategy for enterprise-scale AI workloads. Conversely, if performance lags or costs exceed expectations, it may expose gaps in China's chip ecosystem that will take years to close.