China Unveils Nationwide Supercomputing Network to Bolster Computing Power Capabilities
In the age of artificial intelligence, computing power has emerged as the most critical resource, fueling the engine of the Fourth Industrial Revolution. Consequently, it has become a source of intense competition and another front in the ongoing technological rivalry between China and the United States. Seeking to undermine China’s progress in artificial intelligence, the United States has imposed embargoes on high-performance AI chips, aiming to create a scarcity of computing power. However, mirroring its approach in other domains, China is devising strategies to circumvent this obstacle, undeterred in its pursuit of technological advancement.
For over a decade, China has held the mantle of possessing the world’s largest quantity of supercomputers. Now, leveraging this formidable computing power, the nation’s national supercomputing network has officially gone live, interconnecting numerous supercomputing centers nationwide into an integrated network and service platform for computing power.
An important goal of building the national supercomputing network is to closely match supply and demand through a market-based operation and service system, enabling coordinated scheduling of computing resources. This effort will effectively support original scientific innovation, drive major engineering breakthroughs, boost high-quality economic development, and more, serving as the “highway” for the country’s digital transformation.
Construction of this national supercomputing network officially began in April 2023. Throughout the process, by using an integrated system for computing scheduling, data transmission, and collaboration among different parties, it has created close connections among all industry players involved in computing supply, software development, data trading, model services, and beyond. This has led to a market-driven, internet-based, standardized environment for advanced computing services. To further attract more service providers to join, the “National Supercomputing Network Platform Empowerment Support Program” was officially launched, aimed at driving advances in computing technology and application development.
In building the national supercomputing network, alliance members have played an essential role. As of now, the alliance has 128 members spanning national supercomputing centers, regional computing centers, supercomputer manufacturers, computing operations, network operations, application software, technical services, and more across the computing industry.
Currently, over 200 application, data, model, and other service providers have joined the national supercomputing network platform, offering more than 3,200 products. These cover cutting-edge digital innovation areas such as scientific computing, industrial simulation, AI model training, and more, meeting society’s demand for advanced computing services.
It is hoped that the national supercomputing network will help reduce the current imbalance between supply and demand for computing power, speed up the transformation into intelligent automation, and energizing a data-driven smart economy. In recent years, while China has made remarkable progress in building computing facilities, the rapid advancement of technologies like artificial intelligence has increased the need for even more computing power. Computing centers now urgently need to move beyond the existing single-entity operation model.
To guide the subsequent development and construction of the national supercomputing network, aligning it with China’s long-term strategy, the government also releases the “Supercomputing Internet White Paper”. Drawing on input from supercomputing network alliance experts with a global perspective, the white paper outlines the history and current status of domestic and international supercomputing infrastructure. It explains the operational architecture, operating mechanisms, reference models, and more of the national supercomputing network platform. The white paper also proposes industry development initiatives in areas like technical breakthroughs, application demonstrations, standards creation, talent development, and data sharing for the supercomputing network.