Alibaba's cloud revenue jumps 40% on AI demand

Chinese tech giant Alibaba said on May 13 that its Cloud Intelligence Group’s external revenue rose 40% year-on-year in the March quarter of fiscal year 2026.

This comes as public cloud demand and sales of AI products increased, while group revenue for the quarter and full year each rose 11% on a like-for-like basis.

Cloud revenue reached 41.6 billion yuan (US$6.1 billion), and AI-related products made up 30% of external cloud revenue, according to the company.

Alibaba’s China ecommerce revenue also rose 6% to 122.2 billion yuan (US$18 billion), with customer management revenue up 8% on a like-for-like basis, and its board approved a cash dividend payout of about US$2.5 billion.

🔗 Source: Alizila

🧠 Food for thought

Implications, context, and why it matters.

The strategy behind Alibaba’s cloud surge

Alibaba put more weight on outside revenue after its March 2023 split into six separately run units, including the Cloud Intelligence Group, which was set up to run with more independence 1.
The increase also ties to a push for homegrown hardware in China as geopolitical risks rise 2.
Alibaba’s chip unit, T-Head (Pingtouge), which designs semiconductors, has shipped “several hundred thousand” Zhenwu PPU chips in total. Alibaba Cloud uses them in multiple clusters that serve more than 400 clients 2.
Chinese internet companies are following two tracks. They buy advanced Nvidia chips when they can while also backing domestic AI processors 2.

China’s AI self-sufficiency reshapes the AI compute landscape

Alibaba’s domestic AI chips and AI products are helping build a wider ecosystem that cuts into Nvidia’s lead in China’s AI chip market 2.
International Data Corporation (IDC), a technology market research firm, estimated that Nvidia held about 62% of China’s AI chip-powered server market by value in the first half of 2025, while Chinese chips together reached about 35% 2.
Developers in China are spending more time on compatibility layers, software that helps programs run on domestic alternatives. That can weaken the pull of Nvidia’s CUDA software ecosystem, its platform for programming AI chips 2.
The shift could split the global AI market into a Western-centered technology stack and a separate ecosystem taking shape in China 2.