Key Highlights
- Alibaba partnered with China Telecom to activate a massive 10,000-chip AI computing facility in Shaoguan, located in Guangdong province, utilizing Alibaba’s proprietary Zhenwu semiconductor technology.
- This facility represents the first large-scale implementation of Zhenwu-based computing infrastructure in China’s Greater Bay Area, capable of handling AI model training with parameter counts in the hundreds of billions.
- The facility delivers enhanced performance metrics, including 30% improved training and inference capabilities, with throughput per card achieving nearly tenfold increases over prior technology generations.
- Future expansion plans target scaling the infrastructure to 100,000 chips, while providing on-demand computing services to smaller enterprises through China Telecom’s access framework.
- This deployment comes shortly after a comparable 10,000-chip facility utilizing Huawei Ascend 910C processors became operational in Shenzhen.
Alibaba (BABA) has partnered with China Telecom to bring online a substantial AI computing infrastructure featuring 10,000 chips in Shaoguan, situated in Guangdong province. The entire facility operates on Alibaba’s proprietary Zhenwu AI processors, engineered by its T-Head semiconductor division.
Alibaba Group Holding Limited, BABA
This deployment represents a significant milestone as the inaugural large-scale Zhenwu chip implementation throughout the Greater Bay Area. Alibaba Cloud characterized this initiative as advancing China’s AI computing capabilities “from elite performance achievements toward widespread commercial deployment.”
The infrastructure employs an advanced high-throughput networking framework enabling all 10,000 processors to function cohesively as a unified supercomputing system. Alibaba reports this configuration achieves 30% enhanced training and inference performance, with individual card throughput increasing approximately tenfold compared to legacy architectures.
This computing environment supports training of AI models containing hundreds of billions of parameters — positioning it alongside the most sophisticated AI development platforms worldwide.
The system achieves 4-microsecond latency performance, which Alibaba credits to the interconnected network design linking the processors. This specification proves critical for commercial AI applications requiring rapid response capabilities.
China’s Strategic Investment in Homegrown AI Computing
This launch aligns with comprehensive national initiatives. Beijing incorporated intelligent computing infrastructure into its 15th five-year strategic framework last month, while an August State Council AI directive emphasized optimized development of computing capabilities throughout China.
By mid-2024, China’s aggregate computing capacity reached 962,000 petaflops — representing 21% of global capacity and reflecting 73% annual growth, per data from the China Academy of Information and Communications Technology.
The Shaoguan infrastructure has already been implemented across medical technology and sophisticated manufacturing sectors. Small and medium enterprises can obtain computing resources via China Telecom’s distribution network, with flexible card-based or hourly billing options.
Alibaba has additionally announced intentions to expand the facility from 10,000 to 100,000 chips. This scaling strategy targets reduced operational costs and enhanced resource utilization efficiency.
Industry Context: Huawei and China’s Semiconductor Independence
This activation follows a comparable achievement from the previous month, when China’s inaugural 10,000-card intelligent computing facility — constructed with Huawei’s Ascend 910C processors — commenced operations in Shenzhen.
That infrastructure provides 11,000 petaflops of processing capability and has been integrated with an additional 3,000-petaflop facility launched in 2024. Shanghai is simultaneously developing a 10,000-card computing center through an INESA state-owned subsidiary, designed for compatibility with various domestic processor architectures.
Although Chinese semiconductor technology continues to lag behind Nvidia in standalone chip performance metrics, Beijing’s approach emphasizes large-scale cluster configurations and optimized interconnect technology to narrow the competitive gap.
U.S. export limitations on Nvidia processors have intensified China’s domestic semiconductor development efforts. Alibaba’s T-Head division has emerged as a cornerstone of this strategic initiative, operating alongside Huawei.
BABA shares advanced 7.79% during regular trading on the announcement date, with extended-hours activity adding an additional 0.82% gain on its Hong Kong-listed equity (728-HK).
Get 3 Free Stock Ebooks
Discover top-performing stocks in AI, Crypto, and Technology with expert analysis.
- Top 10 AI Stocks - Leading AI companies
- Top 10 Crypto Stocks - Blockchain leaders
- Top 10 Tech Stocks - Tech giants
