China’s Domestic Computing Power Ecosystem Advances with Talent Synergy

The independent and high-quality development of China’s domestic computing power ecosystem has become a key pillar for fostering new quality productive forces, with talent serving as the core link connecting R&D, scenario application and industrial collaboration. Governments, universities, research institutions and enterprises are joining forces to improve talent cultivation and utilization mechanisms, promoting in-depth integration of the talent chain with the computing power, industrial and innovation chains.

As AI applications permeate various industries, demand for computing power has seen explosive growth, bringing unprecedented opportunities for China’s domestic computing power industry. Liu Liehong, Minister of the National Data Administration, revealed recently that by March this year, China’s daily Token calls had exceeded 140 trillion, a more than 1,000-fold increase from 100 billion at the start of 2024.

Against this backdrop, independent control of the computing power base has become an industry consensus. Industry insiders generally acknowledge that the risk of high-end GPU supply disruptions is a real challenge, making it imperative to accelerate the construction of an independent domestic computing power infrastructure. Guojin Securities pointed out that under the dual pressure of supply and demand, the computing power industry chain is expected to enter an “full-chain inflation” cycle in 2026.

Despite the rapid rise of the domestic computing power industry, a gap remains between industrial development and talent training, with a prominent talent bottleneck. While domestic computing power hardware can support large-model operations, software ecosystem migration and adaptation are more complex, as developers face high costs switching from Nvidia’s dominant CUDA ecosystem to domestic platforms. Universities also focus more on algorithm innovation, lagging in cultivating engineering capabilities needed by the industry.

“As technological barriers in the global computing power field continue to deepen, developing domestic computing power and breaking technological monopolies are key to high-quality development, and talent is the core to achieving this goal,” said Tang Xiaoguang, Director of Huawei China’s Strategy and Business Development Department.

To address the talent shortage, multiple parties are exploring solutions. On April 2, the General Office of the Ministry of Industry and Information Technology issued a notice launching a special campaign to empower SMEs with inclusive computing power, explicitly proposing a computing power talent training program.

Industry-university-research collaboration in talent cultivation is also deepening. On April 15, the 10th Huawei ICT Competition China Finals Challenge kicked off at Huazhong University of Science and Technology. With the theme of “Super Intelligence Integration”, the competition focuses on real industrial scenarios, requiring teams to optimize computing tasks based on the Ascend and Kunpeng platforms, including practical skills like MindSpore 2.3 deployment and YOLOv8 model optimization.

More than 160,000 participants from 1,483 universities joined the competition, with 475 teams advancing to the finals. Feng Dan, Vice President of Huazhong University of Science and Technology, noted the event integrates industrial needs into teaching, helping students enhance practical capabilities. Qian Depei, academician of the Chinese Academy of Sciences, said it reflects the global trend of “super intelligence integration” and promotes the domestic computing power ecosystem.

Huazhong University of Science and Technology’s Dian Team, with 23 years of history, has fostered 837 members through real-project cooperation with Huawei.