Key Highlights
- Alibaba partnered with China Telecom to unveil a 10,000-chip artificial intelligence computing facility in Shaoguan, utilizing proprietary Zhenwu semiconductor technology.
- This facility represents the first large-scale Zhenwu deployment in the Greater Bay Area, capable of training massive AI models with parameter counts in the hundreds of billions.
- The computing center delivers 30% improved efficiency for both training and inference operations, with throughput per card reaching nearly 10 times previous-generation levels.
- Future expansion plans call for scaling the facility to 100,000 chips, while small enterprises gain computing access through China Telecom’s service platform.
- This development comes shortly after a comparable 10,000-chip Huawei Ascend 910C facility commenced operations in Shenzhen.
Alibaba (BABA) has partnered with China Telecom to bring online a massive 10,000-chip artificial intelligence computing facility located in Shaoguan, Guangdong province. The entire cluster operates on Zhenwu AI semiconductors, which were designed internally by Alibaba’s T-Head chip division.
Alibaba Group Holding Limited, BABA
This deployment represents a milestone as the first large-scale implementation of Zhenwu technology in the Greater Bay Area region. Alibaba Cloud characterized the project as advancing China’s AI computing capabilities “from cutting-edge performance achievements to widespread industrial deployment.”
ALIBABA LAUNCHES DATA CENTER WITH 10,000 OF ITS OWN CHIPS AS CHINA RAMPS UP AI PUSH
— First Squawk (@FirstSquawk) April 8, 2026
The facility employs an advanced high-performance networking framework that enables all 10,000 semiconductor units to function as one unified supercomputing system. Alibaba reports this configuration achieves 30% superior efficiency in both training and inference operations, with per-card throughput increasing nearly tenfold compared to earlier generations.
The infrastructure can handle the training of AI models containing hundreds of billions of parameters — placing it among the most powerful AI training systems under development worldwide.
Network latency measures just 4 microseconds, a performance metric Alibaba credits to the sophisticated network design connecting the chips. This low-latency characteristic proves critical for enterprise artificial intelligence applications where response time is paramount.
Beijing’s Domestic AI Computing Strategy
This launch aligns with a comprehensive national initiative. Chinese authorities incorporated intelligent computing infrastructure into the country’s 15th five-year strategic plan last month, while an August State Council AI action blueprint emphasized optimized expansion of computing capabilities nationwide.
As of June last year, China’s aggregate computing capacity reached 962,000 petaflops — representing 21% of global capacity, marking a 73% year-over-year increase, based on data from the China Academy of Information and Communications Technology.
The Shaoguan facility has already found applications in healthcare and advanced manufacturing sectors. Small and medium enterprises can purchase computing time via China Telecom’s platform through card-based or hourly billing models.
Alibaba has also outlined plans to expand the cluster tenfold from 10,000 to 100,000 chips. This scaling initiative targets cost reduction and enhanced overall resource utilization.
Industry Context: Huawei and China’s Semiconductor Independence Drive
This announcement follows closely behind a similar achievement last month, when China’s inaugural 10,000-card intelligent computing cluster — constructed using Huawei’s Ascend 910C processors — became operational in Shenzhen.
That Shenzhen facility provides 11,000 petaflops of computing power and has been integrated with a separate 3,000-petaflop system that launched in 2024. Shanghai is simultaneously developing a 10,000-card cluster through an INESA state-owned subsidiary, engineered for compatibility with various domestic chip architectures.
Although Chinese semiconductors currently lag behind Nvidia in individual chip performance metrics, Beijing’s approach emphasizes large-scale cluster design and optimized networking infrastructure to narrow the competitive gap.
U.S. export controls on Nvidia technology have significantly accelerated China’s domestic semiconductor development efforts. Alibaba’s T-Head division has played a pivotal role in this national push, working alongside Huawei.
BABA shares advanced 7.79% during regular trading on the announcement date, with after-hours activity adding another 0.82% gain on its Hong Kong-listed shares (728-HK).
