In the era of artificial intelligence, China was the first country in the world to put forward the major concept of data as a factor of production. This conceptual innovation elevated data to an unprecedented level of strategic importance.
Deng Zhonghan, “Chip” Journey: China’s Strength in the Computing Power Era, Digital China Summit Organizing Committee, May 16, 2025
The Digital China strategy rests on a core premise: to treat data as a factor of production, China must first build the infrastructure to transform raw data into secure, interoperable, and valuable digital assets.
In a May 16 interview with the Digital China Summit Organizing Committee, Vimicro Chairman and CEO Deng Zhonghan outlined a technical roadmap for advancing the national Digital China strategy through secure data architecture, AI-optimized chips, and scalable compute infrastructure. Deng emphasized a core premise of Digital China: if China intends to treat data as a true factor of production, it must build the technical systems capable of transforming raw data into secure, interoperable, and economically valuable digital assets. His comments align closely with the Party’s broader goal of unlocking the strategic value of data to drive economic modernization and national competitiveness.
Vimicro Corporation, one of China’s leading fabless semiconductor firms specializing in embedded multimedia signal-processing chips, sits at the center of this effort. Beyond his role as CEO, Deng is also an academician of the Chinese Academy of Engineering, chief architect of the Starlight China Chip Project, and a member of the Digital China Summit Organizing Committee. His analysis therefore reflects both technical authority and direct insight into the strategic objectives guiding China’s digital transformation.
A central theme in Deng’s remarks is the decisive importance of compute. Breakthroughs in artificial intelligence depend on algorithms, data, and computing power but compute, he argued, remains China’s binding constraint. Vimicro’s R&D strategy targets next-generation intelligent chips that integrate neural-network processors with real-time audiovisual codec engines, enabling high-throughput, low-latency processing at the edge. These chips are designed for mission-critical domains such as smart transportation, industrial automation, and public security, areas where performance, autonomy, and data sovereignty are non-negotiable.
Deng also highlighted the National Standard GB35114 as a model for China’s emerging data governance architecture. Vimicro is a pioneer in China’s national standard (国标) approach to technology, designed to create secure and controllable (自主可控）ecosystems. GB35114 governs data security in public-security video surveillance networks and incorporates tamper-resistant digital watermarking, cryptographic data signatures, and secure encrypted transmission protocols. These measures support graded data classification, rights confirmation, and full-chain traceability. In Deng’s view, the framework anticipates the challenges of an AI-driven economy in which vast quantities of sensitive public and enterprise data circulate across distributed systems. As such, it provides a template for secure data operations across the entire digital economy.
Vimicro’s work is closely aligned with national industrial priorities under the Ministry of Industry and Information Technology, positioning the company as an essential technical enabler of Digital China. Digital technology, Deng noted, is reshaping the structure of global competition. To capture the full “digital dividends,” China must accelerate implementation of the Three-Year Data Elements X Action Plan (2024–2026), expand AI deployment across industries, and convert raw data into resources, assets, and tradable production factors. In this vision, compute power, secure data architecture, and intelligent chips form the backbone of China’s next stage of digital modernization.
A lightly edited Google Translate machine translation of the published Deng Zhonghan interview follows. An archived copy of the original video interview is also included, transcribed in Chinese and translated into English. I present both the article and the video because they differ slightly. For example, the article prominently mentions Digital China, while the video does not, a notable discrepancy, given that both were produced by the Digital China Summit Organization Committee. The article also misstates the video interview’s original title. I use the correct title in this blog post.
Published: 2025.05.16 | Source: Digital China Summit (DCS) Organizing Committee | 数字峰会组委会
Interviewee: Deng Zhonghan | 访谈人物：邓中翰
Academician, Chinese Academy of Engineering | 中国工程院院士
Chief Architect, “Starlight China Chip Project” | “星光中国芯工程”总指挥
In recent years, China’s digital economy has experienced rapid expansion. The scale and sophistication of its digital infrastructure have improved markedly, digital technologies and industrial systems have matured, and the value of data as a fundamental factor of production has become increasingly salient. | 近年来，我国数字经济快速发展，数字基础设施规模能级大幅跃升，数字技术和产业体系日臻成熟，数据作为关键生产要素的价值日益凸显。
According to Deng Zhonghan, academician of the Chinese Academy of Engineering and chief architect of the “Starlight China Chip Project,” “China was the first country to formally conceptualize data as a factor of production in the era of artificial intelligence. This designation elevated the strategic importance of data to unprecedented levels.” He notes, as a major data power, China must leverage its advantage in data elements and convert data into a factor of production, thereby serving the advancement of New Quality Productive Forces. | “人工智能时代，中国率先在世界上提出‘数据要素’这个重大概念。作为一种生产要素，它把数据的重要性提高到前所未有的高度。” 中国工程院院士、“星光中国芯工程”总指挥邓中翰表示，中国是一个数据大国，我们要充分发挥数据要素的优势，将它转变为生产要素，为新质生产力服务。
Unlock the Value of Data to Advance Digital China Construction | 充分挖掘数据价值，服务数字中国建设
At present, digital technology has become a key force in the restructuring of global factor resources, reshaping the global economic structure, and reconstructing the global competitive landscape. The digital domain is undoubtedly at the forefront of great power competition. | 当前，数字技术已经成为重组全球要素资源、重塑全球经济结构、重构全球竞争格局的关键力量，数字领域无疑是大国竞争的前沿赛场。
Deng explained “in the AI era, governments, enterprises, and other social actors are not only data consumers but also generators of data. To fully capture the dividends of data, we must adopt an open and proactive posture aligned with the Three-Year Data Elements × Action Plan (2024–2026). This policy framework aims to expand the adoption of AI technologies and amplify the multiplier, additive, and compounding effects of data. The ultimate objective is to construct a digital economy where data serves as a critical factor, thereby enabling High Quality Development of Digital China construction.” | “在人工智能时代，政府、企业等社会活动的参与者都是数据的使用者，也是数据的创造者，为更好地享受潜在数据要素价值带给我们的红利，我们应当采取积极开放的态度，按照《“数据要素×”三年行动计划（2024—2026年）》，广泛应用与普及人工智能技术，发挥数据要素的放大、叠加、倍增作用，构建以数据为关键要素的数字经济，推动数字中国建设的高质量发展。”邓中翰院士表示。
Despite China’s extensive data foundation, high level of digitalization, and vast population base, Deng notes that the latent value of its data resources remains largely underutilized. Public data constitutes the largest share but is generally not accessible; enterprise data is only partially shared. To address this, Deng calls for maximizing the latent value of public data and encouraging enterprises to open their data. By transforming data into resources, assets, and tradable factors, China can convert raw data into value-added products and services that support social and industrial development. | 他说，中国人口众多，数字化程度较高，数据基础庞大，然而大量数据的内在价值还没有完全释放出来。公共数据基数最大，几乎不对外开放，企业数据部分开放，我们应该充分挖掘公共数据的潜在价值，同时引导企业开放数据，将数据资源化、资产化和要素化，让数据转变成产品和服务，更好地赋能社会和产业发展。
However, Academician Deng Zhonghan cautioned that in the process of transforming raw data into data resources, and further elevating these into data assets and investing them into society and production, it is essential to leverage artificial intelligence technology. Through technical means as well as standards and norms, a data security mechanism must be established to ensure the security of this data. | 不过，邓中翰院士提醒说，在将原始数据转化为数据资源，再将数据资源上升为数据资产，并投入到社会和生产的过程中，需要借助人工智能的技术，通过技术手段和标准规范建立数据安全机制，确保数据的安全性。
Citing the example of public security video surveillance, Deng references National Standard GB35114 (Technical Requirements for Information Security in Public Security Video Surveillance Networks). This standard employs multi-tiered safeguards such as tamper-proof digital watermarks, video data signatures, and video data encryption to conduct data authentication and protection, ensure security in data transmission and processing during public security video collection, allow data owners and users to use data with greater confidence, and advance the full release of the value of data elements. | 以公共安全视频采集领域为例，按照GB35114《公共安全视频监控联网信息安全技术要求》，通过数据防篡改电子水印、视频数据签名、视频数据加密等分级手段进行数据的确权和保护，实现公共安全视频采集过程中数据传输和处理的安全，让数据拥有者和使用者能够更放心地使用数据，促进数据要素价值的充分释放。|
Chip Computing Power Determines the Level of Intelligence. Vimicro is Increasing R&D to Improve Computing Power | 芯片算力决定智能高度，中星微加大研发提升算力
The development of artificial intelligence is inseparable from algorithms, computing power, and high-quality data. In particular, chips, as the carrier of underlying computing power, have a profound impact on the development of artificial intelligence. In the future, the key factor in the development of artificial intelligence will continue to focus on improving chip computing power. Only through the absolute improvement of chips and computing power can we truly achieve higher levels of artificial intelligence. | 人工智能的发展离不开算法、算力，以及高质量的数据。特别是在算力上，芯片作为底层算力的载体，深度影响着人工智能的发展。未来，人工智能发展的关键因素依然集中在芯片算力的提升上，只有芯片和算力的绝对提升，才能够真正实现更高水平的人工智能。
Deng Zhonghan said that the rapid early development of artificial intelligence was due to the computing power provided by chips. With the continuous evolution, iteration, and development of Moore’s Law, a small chip can integrate hundreds of millions, billions, and even tens of billions of transistors, enabling intelligent computing and forming a new type of computing method known as “neural networks.” For example, GPT-4 has become an extremely large model with 1.8 trillion parameters. It comprehensively and deeply learns all knowledge on the Internet, including data from various fields such as audio and video, to create more content, decisions, and even human-computer interaction. | 邓中翰坦言，人工智能之所以能在早期取得如此快速的发展，是基于芯片所提供的算力。随着摩尔定律不断地演进、迭代和发展，一个小芯片就能集成上亿、几十亿、上百亿颗晶体管，进而实现智能计算，形成被称为“神经网络”的新型计算方式。如，ChatGPT斯沃已经成为拥有1.8万亿个参数的超量大模型，它通过全面、深入地学习互联网上的所有知识，包括视音频等各领域的数据，从而创造出更多的内容、决策，包括人机交互等。
“Currently, over 80% of information transmission is accomplished through images and video. ‘Vimicro’ continues to advance the design and product development of integrated circuit chips, facilitating the further secure release of data elements and serving the country’s sustainable development.” Deng Zhonghan acknowledged that the country has made significant progress in areas such as improving chip computing power and model development, and that scientists are exploring how to forge a path to modernization with Chinese characteristics. As an enterprise under the leadership of the Ministry of Industry and Information Technology, ‘Vimicro’ is deeply exploring the value of data to empower social progress and development. | “目前，80%以上的信息传输都是通过图像和视觉完成，‘中星微’持续推进集成电路芯片的设计和产品的构建，助力数据要素进一步安全释放，服务于国家的可持续发展。”邓中翰坦言，国家在提升芯片算力及模型研制等领域做了大量的工作，科技工作者正在研究如何走出一条中国特色的现代化道路。作为国家工信部领导下的企业，“中星微”深入挖掘数据价值，赋能社会进步与发展。
Reportedly, “Vimicro” has participated in the development of 18 national standards and has twice won first place in the State Scientific and Technological Progress and Wu Wenjun AI Science and Technology Awards. Focusing on national standards, “Vimicro” has been deeply engaged in the intelligent chip field. By embedding cutting-edge AI neural network processors and new-generation AI computing technologies such as audiovisual codecs, Vimicro builds a secure, reliable, and autonomous data application environment. This ensures data security during video capture and transmission, enabling intelligentized services to better contribute to safer production, safer transportation, and safer living. | 据悉，“中星微”曾参与制定18个国家标准，两次荣获“国家科技进步”一等奖及“吴文俊人工智能”一等奖。一直以来，围绕国家标准，“中星微”深耕智能芯片领域，通过嵌入最先进的人工智能神经网络处理器，以及相关视音频处理的编解码等新一代人工智能计算技术，构建安全可靠、自主可控的数据运用环境，实现包括视频采集、传输过程中的数据安全，更好地让智能化服务于安全生产、安全交通、安全生活等。
Academician Deng said that “the field of chip research and development also faces many challenges and opportunities. By continuously increasing research and development efforts, I believe that consumers will be able to truly feel the importance of data elements in the era of artificial intelligence in serving work, life and other aspects.” | 在芯片研发领域同样面临着诸多挑战与机遇，通过持续加大研发力度，相信广大消费者一定能真真切切感受到人工智能时代数据要素在服务工作、生活等方面的重要意义。”邓中翰院士如是表示。
An archived copy of Deng Zhonghan’s video interview, along with a Chinese transcript and lightly edited Google machine translation, follows below:
Deng Zhonghan, “Chip” Journey: China’s Strength in the Computing Power Era, Digital China Summit Organizing Committee, May 16, 2025 | 邓中翰, “芯”征程: 算力时代的中国力量, 数字峰会组委会, May 16, 2025
Timestamp: 00:01
Deng: In the era of artificial intelligence, China was the first country in the world to put forward the major concept of data as a factor of production. This conceptual innovation elevated data to an unprecedented level of strategic importance. China can be regarded as the largest generator of data globally, not only because of its vast population, but also due to its relatively advanced level of digitalization across infrastructure and other domains. | 我们国家在这个人工智能时代，率先的在世界上提出了这个数据要素这样的一个重大的概念。作为一种生产要素，把数据提高了前所未有的重要性上。我们国家应该说是数据量产生最大的国家。不仅因为我们的这个人口众多，同时也因为我们的基础设施等各个方面，数字化程度是比较高的。
Interviewer: “How to build a robust data security system?” | 如何构建一个完善的 数据安全体系?
Deng: To transform data into a productive resource, first as a resource, then as an asset, and ultimately data as a factor of production, a sound security architecture is indispensable. China has already begun exploring mechanisms in this domain. For example, in the field of public security video surveillance data, the application of national standard GB/T 35114 enables data classification and certification through a three-tiered grading system. | 数据如何变成资源，变成资产，再变成要素。那么在这过程中，安全机制是个非常重要的一个手段，那目前我们在这方面已经探索在比如说像公共安全视频采集的这些数据里面 如何通过国家标准35114，分三个等级来进行数据的认证。
Timestamp: 01:01
Deng: This graded approach involves digital watermarking to ensure data provenance and prevent tampering, encryption for secure transmission and storage, and legal and technical mechanisms for establishing data ownership and rights. Such layered protections not only safeguard privacy and information security, but also enable secure, autonomous, and controllable systems that empower data holders and users alike. Under this framework of graded protection, valuable data can be more effectively processed and its economic and strategic value as a factor of production more fully unlocked. | 数据的防篡改的电子水印的确权，以及通过密码来进行加密传输和存储处理等。这样三个级别，对数据进行确权和保护，能够更好的实现，包括个人的隐私的安全以及这个信息安全等相关的这些自主可控的这样的方式，来使得呃我们这些数据的拥有者和使用者更好的在这种分级保护之下。
Timestamp: 01:36
Deng: Being able to process this valuable data allows us to better unlock the value of data as a key resource. In the age of artificial intelligence, as we all know, we need algorithms and computing power. Especially regarding computing power, integrated circuit chips, as the fundamental providers of computing power, have consistently been a bottleneck in the development of artificial intelligence. However, significant progress has been made in this area in recent years. | 能够呃对这些有价值的数据进行处理，更好地释放数据要素的它的价值。在人工智能时代，呃，我们大家都知道，我们需要算法，啊，算力，特别是在这个算力上，集成电路芯片作为最底层的算力的提供一直成为这个人工智能发展的一个瓶颈，那么之所以在过去这些年啊取得快速的发展。
Timestamp: 02:06
Interviewer: Which industries and sectors merit priority attention in the future? | 哪些产业和领域是未来可以重点关注的？
Deng: The rapid progress seen in recent years has largely stemmed from continuous iterations and innovations in the semiconductor field, in accordance with Moore’s Law. This has led to the emergence of new computational architectures, most notably neural networks. Looking ahead, the future of AI will continue to hinge on breakthroughs in chips and computing power. Only through absolute improvements in both can we achieve higher levels of artificial intelligence. | 在人工智能时代， 我打家都知道我们需要算法，算力，特别是在这个算力上集成电路芯片作为最底层的算力的提供。 一直成为这个人工智能发展的一个瓶颈。 那么之所以在过去这些年去的快速的发展。还是基于芯片领域随着摩尔定律不断地这个引进、迭代和发展，逐渐形成了，我们称之为神经网络的这样的一种计算的方式。未来人工智能最关键的依然是在芯片和集成电路的算力的提升上，只有[芯片]和[算力]的绝对的提升，我们才能够真正地实现更高水平的人工智能。
Timestamp: 02:38
Deng: History and forward-looking technological trends confirm this as an ironclad principle. China has invested considerable effort in this direction. As Chinese scientists, we are committed to exploring a path of modernization with Chinese characteristics. Vimicro Corporation (中星微), which has twice received the National Science and Technology Progress Award and most recently the Wu Wenjun AI Science & Technology Award, is determined to contribute to next-generation AI chip technologies in this new era. | 从过去的历史看，和对未来的这个技术展望上看。这是一个铁的规律，我们国家也做了很多的努力工作，我们科技工作者正在研究如何走一条中国特色的现代化的道路。那作为中星微，曾经荣获过两次“国家科技进步一等奖，” 去年呢，刚刚又荣获了“吴文俊人工智能一等奖”的团队，一定会要在这个时代，为人工智能的芯片技术做出一些新的贡献。