IDC Projects China AI Data Infrastructure to Hit $154.8 Billion by 2035, Opening New Space for Domestic Databases
IDC forecasts China's AI data infrastructure market will grow at a 37.7% CAGR from 2025 to 2030, reaching $73.8 billion by 2030 and $154.8 billion by 2035, as data capability becomes the main bottleneck for AI value and domestic databases expand into AI data platforms.
IDC estimates the China AI data infrastructure market at about $14.9 billion in 2025. The firm divides the sector into AI data platforms, AI data intelligence and AI data services. AI data platforms handle unified management and processing of multimodal data and serve as the data foundation for AI applications. AI data intelligence uses AI to improve data governance, analysis and usage efficiency. AI data services extend from traditional data labeling to knowledge base operations, evaluation data, synthetic data and continuous services. AI data intelligence is expected to be the fastest-growing segment, with a 48.2% CAGR from 2025 to 2030.
Wang said that in most enterprise business scenarios, large models' capabilities are already basically sufficient. The obstacle to producing business value from AI is shifting to data capability. IDC expects China's data generation volume to rise from 76.05 zettabytes in 2025 to 146.65 zettabytes in 2030. Documents, images and other unstructured data, along with vector data, are increasingly entering enterprise data assets. Traditional data architectures mainly serve transactions, analytics and reporting, while generative AI and agents require data to be stored, unified, retrieved in real time, understood and invoked, with security and compliance maintained throughout. AI is pushing data infrastructure from supporting business to serving intelligence.
IDC said future competition in AI data infrastructure will depend on who can first form a complete loop covering data management, data intelligence, data services and agent consumption. Wang pointed to a path of combining small models with multimodal foundations, integrating industry knowledge and scenario data, and using small-model fine-tuning to achieve efficient, controllable production-grade AI deployment. The forecast treats AI data infrastructure not as a market appearing from nowhere, but as an overall upgrade of traditional data infrastructure. Databases, data warehouses, data lakes and data governance capabilities must adapt to AI applications and agents. Data flows created by AI inference, intermediate data and agent calls will further expand demand for data management and services.
OceanBase is one of the domestic database vendors moving in this direction. While continuing to expand in distributed databases, OceanBase is extending toward an AI data platform. Through capabilities such as Lakebase, it is integrating databases, open storage, analytics and multimodal data processing to build a unified data foundation for structured, semi-structured, unstructured and vector data. Through DataPilot and other Data Agent capabilities, it allows AI to participate in data queries, analysis and usage, extending data capabilities to business users and agents.
For OceanBase, AI is not a separate business line outside its traditional database work. The company is positioning its existing data infrastructure capabilities for a larger market, moving from serving core enterprise data to becoming a unified foundation for all enterprise data and a data foundation for agents. For domestic databases, AI brings a market opportunity larger than the traditional database market.