Western: Third-party IDC service providers' market share is expected to increase, accelerating the industry transformation.

Western released a research report stating that IDC mainly focuses on long-term hosting and leasing, with a long investment return period, while AIDC has a shorter investment return period. IDC demand is shifting towards AI, with an increasing demand scale and a structural evolution towards customization and diversification. Operators dominate the market, but third-party IDC service providers are expected to gain market share due to their efficiency and flexibility advantages. With the wave of AI and the promotion of the "East Calculation, West Data" policy, the IDC industry is accelerating its evolution from traditional data storage and network service platforms towards computing-intensive, green energy-saving, and intelligent directions. The main points of the Western report are as follows: What kind of business is IDC? IDC provides specialized physical infrastructure services for enterprises. The overall construction of a data center involves planning, design, construction, testing, and operation. Taking Guangdong Aofei Data Technology as an example, the entire IDC project cycle from inception to full operation is approximately 12 months, with an investment return period of about 6.7 years and an IRR of 12.09%-12.30%. What are the differences between AIDC and IDC? 1) Technologically, IDC mainly uses general-purpose servers with low cabinet power density and relies on air cooling; AIDC mainly uses high-density AI servers with liquid cooling technology. 2) Business-wise, IDC focuses on general computing needs and basic data storage, with a long investment return period through long-term hosting and leasing; AIDC specializes in AI training, inference, and high-performance computing, offering vertical and specialized services such as customized cabinet deployment and dedicated cooling solutions, with a shorter investment return period. How to understand IDC demand? 1) In terms of demand drivers, the focus is shifting from cloud computing to AI. 2) In terms of demand scale, construction efforts continue to expand. 3) In terms of demand structure, major Internet companies are still the foundation, but demand from midsize clients, traditional enterprises, and foreign clients is growing rapidly, driving IDC services towards customization and diversification. How to grasp the evolution of the supply side? 1) By 2024, the standard number of active rack units will reach 8.8 million, with a continuously improving average deployment rate. In terms of supply structure, first-tier cities and their surrounding areas are the core IDC regions, with demand spillover; the western regions are gradually rising. 2) The core barriers on the supply side are resource endowment, delivery speed, and operational capabilities: the value of resources is strongly related to the geographic location of the IDC, delivery speed affects the ability to secure orders, and post-installation operational capabilities have a significant impact on equipment safety, energy efficiency, and stability. Additionally, under new trends, the emergence of new formats such as computing power leasing can impact IDC order acquisition. 3) In terms of competitive landscape, operators dominate the market, but third-party IDC service providers have the potential to increase market share due to their efficiency and flexibility advantages. Where is IDC heading? 1) Technological architecture transition: ultra-high density and heterogeneous computing drive the computing power revolution. 2) Green and efficient cooperation: low-carbon power supply and liquid cooling reshape energy efficiency standards. 3) Trend towards AIDC cluster development: global resource optimization under the "East Calculation, West Data" strategy. 4) Improving delivery speed of AIDC, with prefabricated and modular construction becoming an industry trend. 5) Globalization of AIDC construction, with Southeast Asia becoming a core region for domestic AIDC construction companies to expand overseas. Risk warning: AI development falls short of expectations; policy risks; increasing industry competition.