China's science and technology race moves at the pace of a decade each year

Baek Seo-in

The author is a professor at Hanyang University

China's science and technology sector continues to generate headlines at a dizzying pace. One year after my previous visit, I returned to see how much had changed. As competition for AI leadership intensifies, Korea has also worked hard over the past year. But how far has China moved during the same period?

My first stop was Huawei's research and development campus in Shanghai. Most of the construction that had been underway during my last visit was now complete, making the campus even more impressive. During the briefing, one statistic immediately stood out: 30,000 researchers who had an average age of 31.4.

Compared to a year earlier, Huawei had hired 10,000 more researchers while maintaining the same average age. In effect, the Shanghai campus expanded its research work force by 50 percent in just one year.

Another display showcased Huawei's AI data center and cloud infrastructure, built entirely with its own technologies, from power systems to chips. More striking was a benchmark comparing token processing output time between Huawei's neural processing units and Nvidia's H20 processors. Although Huawei conducted the test itself, its chips processed tokens an average of 2.5 to four times faster. That also explained why Nvidia's China-bound GPUs reportedly failed to become a meaningful bargaining chip during last month's U.S.-China summit. China already possesses competitive domestic alternatives.

The same acceleration was evident in robotics.

Last year, Chinese humanoid robot makers drew global attention with robot marathons and other elaborate demonstrations. This year, they had moved beyond spectacle and into factories.

At AgiBot, hundreds of robots and workers were conducting field tests inside a warehouse-sized facility. Supported by a work force in which roughly 80 percent are engaged in research and development, the company now upgrades its robots every few months. After selling about 5,000 units last year, it aims to ship 40,000 this year.

A two-hour flight away in Shenzhen, Ubtech was expanding its deployment of humanoid robots in manufacturing and logistics. The company has already supplied about 1,000 robots to production lines at Zeekr, Boeing and SF Express. An automatic battery-swapping system now allows continuous operation, and future models are being designed to lift heavier loads and work faster. Company representatives said that they are gradually replacing Nvidia chips with processors from Huawei and Alibaba to reduce dependence on U.S. technology.

At the Institute of Quantum Information and Quantum Technology Innovation under the Chinese Academy of Sciences, I met a young researcher who had completed postdoctoral work in the United States before returning to China. A graduate of Tsinghua University's elite Yao Class, he studied under Prof. Kim Ki-hwan, who recently returned to Korea, and later worked with Duke University Prof. Jungsang Kim.

After returning home, the graduate was immediately appointed as a tenured professor at the academy. While his former adviser in Korea is still establishing laboratories and seeking research funding, the graduate received the equipment and budget to operate three laboratories in Shanghai from the moment he was hired. He is also preparing two more laboratories at the academy's main campus. He now leads research on both ion-trap quantum communication and quantum computing and works alongside Pan Jianwei, widely regarded as the father of China's quantum program. Their discussions already focus on how to recruit 1,000 principal investigators to the institute.

As a scholar who studies China's technological innovation and advises both government and industry, I visit leading researchers, companies and research centers as often as possible. I recognize that they naturally emphasize their strengths while downplaying weaknesses, and I remain conscious of the limitations of China's system, as well as my own biases.

Even so, every visit leads me to the same conclusion. China is innovating with extraordinary speed and determination. Acknowledging that fact does not mean denying China's weaknesses or Korea's opportunities. It means recognizing that our competitor is advancing much faster than many of us assume. While China has spent the past year progressing as though it were a decade, what have we accomplished?

We should stop comforting ourselves with the belief that tougher U.S. pressure will inevitably weaken China, that Chinese technology will be shut out of global markets because of security concerns or that Korea will naturally replace it as the trusted alternative. Korea's remaining edge lies primarily in memory semiconductors, and even that advantage cannot be taken for granted.

Nvidia CEO Jensen Huang may have visited Korea for highly publicized events, but he travels to China repeatedly to meet companies such as AgiBot. At the Chinese Academy of Sciences' quantum institute, the walls are covered with the signatures of Nobel laureates in physics who have visited. We should not celebrate ambitious plans before they become a reality. Beyond slogans, only sustained innovation will determine who remains competitive.