Alibaba has launched a new data center powered by its own chip technology, a major step in China’s effort to build a self-sufficient AI sector.
Located in Guangdong province and developed in partnership with China Telecom, the data center is equipped with 10,000 proprietary AI chips designed by Alibaba. These processors are built to handle both the training of large-scale AI models and the execution of inference tasks, supporting systems with hundreds of billions of parameters.
The companies indicated that the facility could expand tenfold in the coming years, demonstrating the scale of China’s infrastructure ambitions. Alibaba, through its semiconductor unit, has spent several years developing in-house chips while simultaneously expanding its cloud computing business, which remains one of its fastest-growing segments.
Working Around US Export Controls
Since 2022, U.S. export controls have limited China’s access to advanced chips from companies such as NVIDIA, prompting a surge in domestic development efforts. In response, Chinese firms have accelerated investments in homegrown hardware capable of supporting increasingly complex AI workloads.
Alibaba’s new data center also reflects a distinct approach to AI investment compared with Western counterparts. While U.S. tech companies are expected to spend hundreds of billions of dollars on AI infrastructure this year, Chinese firms have taken a more targeted path. Their investments are often directed toward applications with clear commercial outcomes, including healthcare and industrial systems.
Beyond its data center launch, Alibaba is also reorganizing internally to sharpen its AI focus. The company recently formed a new technology committee led by CEO Eddie Wu, bringing together senior technical leadership to streamline development efforts.
The partnership with China Telecom adds significant growth potential. As a major state-backed operator with nationwide infrastructure, China Telecom is well positioned to scale AI services and distribute computing capacity across industries. The collaboration points to a model in which telecom providers and cloud companies jointly build and monetize AI infrastructure.
Compete Globally?
Questions remain about how effectively domestic chips can compete with established alternatives. While Alibaba’s processors are designed for AI-specific workloads, matching the performance and efficiency of leading Western products remains a complex challenge. Developers often rely on mature programming frameworks and tools that have evolved over decades, creating an additional hurdle for newer entrants.
Even so, the Alibaba facility may serve as a proving ground. By deploying its chips at scale, Alibaba can gather data and iterate on future designs to refine performance. This feedback loop is critical in a sector where real-world workloads often reveal limitations not apparent in planning documents.
The larger implication is the possible divergence in global AI infrastructure. As China builds out its domestic tech sector and the U.S. continues to lead in advanced chip design, the industry may evolve into parallel ecosystems with distinct technologies and standards. For companies operating within China, the success of initiatives like Alibaba’s will likely determine the pace and direction of that transition.
