According to a research report released by Sealand Securities, AI has emerged as the core engine driving growth in the cloud computing industry. The massive computing power demands generated by training and inference of large models, along with the wide-ranging AI applications that follow, are significantly raising the ceiling of the cloud computing market. To seize this opportunity, major global cloud service providers are increasing capital expenditures for high-performance infrastructure development while simultaneously enhancing AI development tools and platform services, expediting their large-scale implementation across various industries. Based on the incremental space and profit prospects brought by AI, the recommendation rating for the computer industry is maintained. Key points from Sealand Securities include: The transition of cloud computing to intelligence and industry space reassessment.
IaaS, or Infrastructure as a Service, offered by cloud service providers, encompasses three categories of IT basic resources: computing, storage, and networking. Together with PaaS (Platform as a Service) and SaaS (Software as a Service), they form the main product model for cloud service providers. Cloud service providers operate in the midstream of the industry chain, providing various core cloud service products; upstream resources include chips and IT infrastructure such as CPUs, GPUs, ASICs, servers, storage devices, routers, and switches; downstream encompasses applications and end users, covering enterprise users (finance, healthcare, manufacturing, etc.), individual users (cloud storage, online office), and developers. The cloud computing industry has undergone phases of resource cloudification (2006-2013), cloud-native development (2013-2020), and ubiquitous computing power (2020-2023), with 2023 seeing the large model training wave triggered by ChatGPT, entering the intelligence phase; moreover, AIaaS has technically restructured the three fundamental resources of computing, storage, and networking.
In terms of industry space, the global IaaS market is experiencing rising growth. According to iResearch, the domestic IaaS market is expected to approach 800 billion yuan by 2029; currently, the cloud computing market remains highly concentrated, with the top four giants competing in the intelligent cloud positioning.
The top global cloud service providers have seen acceleration in AI business growth, driving increased capital expenditures. Before 2022, the revenue growth rate of the global top four cloud companies was continuously slowing, as traditional cloud services reached maturity. However, following the large model technology and AI commercial applications that scaled up at the end of 2022, revenue growth rates for leading cloud providers have stabilized and begun to rise, entering a new growth cycle. Among them, according to CCID Consulting and Changshi Capital, Microsoft's AI services contributed 16 percentage points to Azure's growth in Q2 2025, while Azure AI Foundry handled 500 trillion tokens in the 2025 fiscal year, marking over a 7-fold year-on-year increase; Alibaba Cloud AI-related businesses have maintained triple-digit growth for eight consecutive quarters, becoming a core highlight. In Q2 2025, revenues from AI business for the global top four cloud service providers grew rapidly, with orders continuously increasing over nearly three years. Leading cloud companies are consistently expanding their capital expenditures, shifting investment focus from traditional general-purpose data centers to intelligent computing centers and AI chip investments: Amazon, Microsoft, Google, and Alibaba have significantly ramped up AI infrastructure investment recently, raising their capital expenditure plans for 2025 and beyond, firmly committing to AI transformation.
Looking ahead, the inference market will become a new driving force, with commercial opportunities shifting towards the application layer. The technical iteration direction for large models focuses on enhanced inference, multi-modality, agent technology, and long context, which will strengthen the practicality and accuracy of AI in more complex and “productive” key scenarios, accelerating the deployment of AI applications and shifting the focus of the large model industry development from training to inference.
The "Matthew Effect" in the cloud industry is becoming increasingly evident, as core underlying resources such as data, computing power, and storage are accelerating their concentration towards a few leading providers like Amazon, Microsoft, Google, and Alibaba Cloud, leading to a solidified market landscape. At the same time, the value height of cloud services is shifting from the infrastructure level (IaaS) to the platform level (PaaS) and model-as-a-service level (MaaS), even further up to the SaaS level; especially, AI-native applications and industry solutions have emerged as new growth engines. Currently, leading cloud providers generally perceive a tightening supply and demand situation and see clear acceleration in revenue growth ahead; the firm believes that the current acceleration of capital expenditures and revenue among cloud providers has resonated, and the new AI cloud cycle may just be beginning.
Risk Warning: Intensified U.S.-China rivalry, downstream industry demand recovery not meeting expectations, AI large model development not reaching expectations, technology progression not meeting expectations, increased market competition, exchange rate fluctuation risks, and performance of relevant companies falling short of expectations.
Comments