China's roadmap to self-sufficiency in storage and neutralization of foreign risk - Analysis

China's Anti-Vulnerability Plan to Secure AI

The China is the world's largest producer of data, yet this abundance conceals a strategic Achilles' heel: its inefficient management and preservation. Despite the stratospheric volumes of information generated, a significant amount of this critical "raw material" for Artificial Intelligence (AI) is lost or is not stored and processed efficiently. This phenomenon, described as "data that should be stored is not," represents a systemic gap that risks slowing China's ambition to surpass its rivals in AI, as advanced models require datasets of the highest quality and immediate accessibility for training and technological development.

GettyImages

It is precisely in response to this structural vulnerability that the Research Centers in Information and Communication Technologies (ICT) are mobilized to generate the most ambitious national strategies. It is within this perspective that the Research Report on Advanced Storage Centers (2025) is placed. Drafted by the Institute of Cloud Computing and Big Data (云计算与大数据研究所), an arm of the China Academy of Information and Communications Technology (CAICT - 中国信息通信研究院), this document is not a mere technical white paper. It represents an official roadmap that translates China's global leadership aspirations—particularly in Artificial Intelligence—into a concrete infrastructural plan, directly addressing its vulnerabilities and technological dependence. The following analysis breaks down this strategy in terms of intelligence, highlighting how China intends to transform advanced storage into a pillar of its national security and self-sufficiency.

To quickly bridge this gap and reduce vulnerability to foreign technological dependence, Beijing has launched the project for the Advanced Storage Centers (ASC - 先进存力中心). These centers are conceived as the nation's new critical information infrastructure, essential alongside computing power and communication networks. Their objective is twofold: to transform the raw quantity of Chinese data into secure, efficient knowledge assets ready for algorithmic training, and at the same time to serve as the primary tool for achieving the goal of technological self-sufficiency (科技自立自强) set out in the 14th Five-Year Plan (14th FYP). The ASC adopts a "1+4+N" architecture (架构体系): an advanced storage base that provides four fundamental service capabilities (Custody, Governance, Circulation, and Data Development) to power a multitude of critical application scenarios. This is China's plan to win the "data war" through resilience and data industrialization.

Strategic context: The need for an intelligence infrastructure

China has recognized that superiority in Artificial Intelligence rests on three pillars: computing, algorithms, and data. While computing and chips (such as advanced semiconductors) remain its greatest point of vulnerability due to export restrictions imposed by the United States, data represents its primary and irreplaceable advantage in terms of volume.

The problem is that this volume has translated into inefficient management. The report highlights that the growth of storage capacity (estimated at +20.81%) is not keeping pace with the growth of data production (+24.99%), with the storage rate dropping to 2.80%. This inefficiency is not just a technical problem; it is a strategic intelligence risk because the loss or mismanagement of this data prevents the creation of high-quality datasets, vital for the sectoral fine-tuning of Chinese Large Language Models (LLMs). The ASC is, therefore, a move for data capitalization on a national scale.

The achilles' heel and the 14th five-year plan's remediation strategy

The CAICT analysis fits directly into the logic of the 14th Five-Year Plan (14th FYP) (2021-2025) and the mission of technological self-sufficiency. The 14th FYP identified computing infrastructure as a key sector, aiming for well-defined standards and capacities, including an increase in storage capacity and the share of advanced storage.

The technological gap and the risk of dependence

For decades, China has relied on foreign technologies for critical components of its advanced storage, particularly for controllers, proprietary firmware, and, indirectly, for memory chip components. This dependence is no longer sustainable in a context of trade and technological warfare. The interruption of the supply chain, for example of proprietary software or crucial semiconductor components (such as those needed for All-Flash drives or large-scale data center systems), would represent a catastrophic risk to national security, paralyzing industry, finance, and defense networks.

The countermeasure: Independent innovation and integration

To bridge this gap and eliminate the risk of interruption, the ASC does not merely call for more space but demands a revolution in technological architecture:

• All-flash architecture (全闪存储架构): The massive push for the adoption of All-Flash architecture (solid-state memory) is a signal of the will to develop and produce these high-performance components internally, which are less vulnerable to computation delays and more energy-efficient.

• Indigenous software development: The need for functionalities such as data governance (数据治理), trusted custody (数据可信托管), and the trusted data space (可信数据空间) cannot be met with foreign off-the-shelf solutions, which could contain backdoors or be remotely disabled. The plan pushes for the development of entirely indigenous software and security protocols.

• Storage-compute synergy (存算协同): China cannot afford to waste the power of its chips, which are often less advanced or less numerous than rivals'. The ASC solves this problem by providing computing centers with a high-speed, pre-optimized data flow (performing data format conversion before the computation task), maximizing the utilization rate (利用率) of available computing resources.

Architecture and operational capacities for self-sufficiency

The "1+4+N" architecture is the operational model for self-sufficiency:

1. Trusted custody (数据可信托管 - Shùjù Kěxìn Tuōguǎn): Ensures the efficient accumulation of multi-source and multi-modal data on a national platform. The goal is to guarantee physical and logical reliability through encryption (加密) and solutions against attacks, reducing management difficulty.

2. Data governance (数据治理 - Shùjù Zhìlǐ): The critical process of standardization, classification (分级归类), cleansing (清洗), and cataloging to transform fragmented and raw data into usable data assets (数据资产). This is vital for AI, ensuring higher quality datasets for national model training.

3. Data supply and circulation (数据供给与流通 - Shùjù Gōngjǐ Yǔ Liútōng): Promotes the secure circulation of data between different actors through the Trusted data space (可信数据空间), a technical environment that allows data utilization and exchange while ensuring that the raw data does not leave the platform (the principle of "available but not visible"). This is the foundation for the future unified national data market (全国一体化数据市场).

4. Data development and processing (数据开发与加工 - Shùjù Kāifā Yǔ Jiāgōng): Consists of processing raw data "staying local" (数据留在本地), converting it into reusable data services and intelligence. This principle of "data staying local" maximizes the security and effectiveness of sectoral AI fine-tuning without exposing sensitive information.

Technologies for intrinsic security (内生安全) and resilience

Self-sufficiency is not only productive capacity but also resilience and intrinsic security. The key technologies of the ASC reflect this "native" security principle:

• Inherent storage security (数据存储内生安全 - Shùjù Cúnchǔ Nèishēng Ānquán): Defenses are integrated into the architecture at the hardware and firmware levels. This includes full-chain encryption (全链路加密) and key lifecycle management, ensuring that storage is auditable (可审计), irreversible (不可逆), and zero-loss (零泄露). By building this "data fortress" from the ground up, China aims to isolate its critical infrastructures from potential vulnerabilities implanted or exploitable by foreign state actors through supply chains.

• AI data lake storage (AI数据湖存储技术 - AI Shùjù Hú Cúnchǔ Jìshù): Unifies the management of all types of data (structured, unstructured) and automates metadata capture (元数据自动捕获) and lineage recording (data lineage).

The report refers to the support of important institutions and universities. The CAICT, a research institute under the Ministry of Industry and Information Technology (MIIT), recommends integrating research from universities and research institutes (高校院所) with technology companies (企业) to overcome bottlenecks and promote collaborative innovation.

Power projection and critical application scenarios

The application scenarios of the ASC are the spearhead sectors of China's modernization and global competition strategy:

• Intelligent connected vehicles (智能网联汽车): The ASC is responsible for managing the enormous volume of data generated by autonomous vehicles, essential for fine-tuning driving algorithms and creating "digital twins of the vehicle".

• Authorized operation of public data (公共数据授权运营): The ASC is the central hub for the secure custody and circulation of public data, implementing the model of state data governance.

• Intelligent manufacturing (智能制造): The ASC supports the creation of massive enterprise knowledge bases, transforming unstructured data (such as CAD drawings or production logs) into reusable knowledge resources for AI, crucial for the quality leap of the manufacturing industry.

The Advanced Storage Center is, ultimately, the mechanism through which China intends to remedy its primary infrastructural weakness and project its model of digital sovereignty globally. By establishing these standards and capacities at the national level, Beijing lays the foundation for self-sufficiency and global technological influence in the coming decade.