China formalises 2030 semiconductor self-reliance strategy; Korean firms face benefits and threats

China assigned about 1.77 quadrillion won at this year’s “Two Sessions” to build semiconductor and AI infrastructure, launching in earnest a national strategy aimed at 2030. It is a declaration of a long-term campaign to complete a “technology rise” beyond “technology self-reliance”. For South Korea’s chip industry, a dual structure has opened at once: near-term demand support and medium- to long-term market encroachment.

China formalised at this year’s Two Sessions (the National People’s Congress and the Chinese People’s Political Consultative Conference) a strategy to directly foster semiconductors and AI with state capital. It plans to issue 1.3 trillion yuan in ultra-long special bonds and invest 7 trillion yuan in power grids and AI computing infrastructure alone. An assessment is emerging that it has formalised a long-term posture to complete a “technology rise” beyond “technology self-reliance” against Western export controls.

Premier Li Qiang (李強) presented a 4.5 to 5 percent range for this year’s economic growth target through the government work report at the fourth session of the National People’s Congress. It is groundwork for a long-term campaign to prepare for geopolitical uncertainty such as additional U.S. export controls. Hanwha Investment & Securities said, “Setting a realistic growth target suggests policy capacity will focus on securing medium- to long-term growth engines rather than short-term growth,” and “while property stimulus and low-priced exports centred on price competition weaken, policy capacity will be concentrated on strengthening advanced industries (semiconductors, AI).”

Above all, the key issue at this year’s Two Sessions is securing fiscal resources. China raised the size of general public budget spending for the first time to 30 trillion yuan and additionally set 1.3 trillion yuan in ultra-long special bonds and 4.4 trillion yuan in local government special bonds. Of that, 800 billion yuan will be directly allocated to major national strategic projects. It also specified a target of increasing research and development spending by more than 7 percent a year on average during the 15th five-year plan (2026-2030).

Already last year China’s total social R&D investment exceeded 3.92 trillion yuan, reaching 2.8 percent of GDP. Investment in basic research also exceeded 7 percent of total R&D spending for the first time. Wang Guochen of Taiwan’s Chung-Hua Institution for Economic Research (CIER) said, “By issuing 1.3 trillion yuan in ultra-long government bonds, Beijing is operating on the premise that only when breakthroughs in advanced technologies such as semiconductors are achieved will it create the economic capacity to repay massive national debt.”

The government work report also specified fostering “six emerging pillar industries” that bundle integrated circuits, aerospace, biomedicine and the low-altitude economy, among others. The Chinese government presented a roadmap to grow these six industries, including semiconductors and intelligent robots, into a “10 trillion yuan” strategic market.

Li Lecheng (李樂成), minister of industry and information technology, said at a news conference, “We will vigorously promote a two-way sprint between AI and manufacturing,” adding that “last year China’s core AI industry exceeded 1.2 trillion yuan in size and the number of companies surpassed 6,200.” China will also foster future industries in parallel, including quantum technology, embodied intelligence, brain-computer interfaces (BCI) and 6G. The five-year plan also includes a target of raising the value added of core industries in the digital economy to 12.5 percent of GDP.

In-house semiconductor production push, AI infrastructure build speeds up… “AI and manufacturing, two-way sprint”

China plans to invest more than 7 trillion yuan this year alone in building power grids and AI computing infrastructure. The outline draft of the 15th five-year plan specifies “strengthening the efficient supply of computing power, algorithms and data.” According to Xinhua, the Chinese government’s mouthpiece, China’s smart computing scale has already exceeded 1,590 EFLOPS and it holds 60 percent of global AI patents. The investment is intended to support the success of low-cost, high-efficiency AI models such as DeepSeek with state-led hardware infrastructure.

Nurturing achievements of 504 smart factories and 1,260 5G factories were also disclosed at the National People’s Congress. Kyle Chan, a researcher at the U.S. Brookings Institution, assessed this as “a national push for self-reliance arising from frustration and urgency to break dependence on U.S.-made semiconductors and core technologies.” The Financial Times also said it was “a strong policy intention to promise market opening to foreign-invested companies while thoroughly confining the underlying infrastructure and the ecosystem of the six emerging industries within China through localisation.”

South Korea’s chip industry has reached a crossroads from this year’s Two Sessions in China. As China expands AI computing infrastructure on a large scale, demand for advanced memory, including high bandwidth memory (HBM), could be maintained for the time being. But if the six emerging pillar industries policy gains speed, integrated-circuit self-reliance could erode the China market share of South Korean companies in the medium to long term, creating a structure in which benefits and threats coexist.

The key is the pace of semiconductor self-reliance. Depending on what stage China’s integrated-circuit technology reaches by 2030, the structure of South Korea’s semiconductor exports to China could change fundamentally. An industry official said, “China’s speed is different from ours,” adding, “Considering China’s roadmap through 2030, there is not much time left for a strategic response for South Korea’s semiconductor materials, parts and equipment.”

Seok Dae-geon
2026-03-12