Chinese ASML Emergence, U.S. Semiconductor Jones Act Loom Expert predicts Chinese ASML within a decade and potential U.S. regulations reshaping global semiconductor dynamics
www.chosun.com
Opinion
Chinese ASML Emergence, U.S. Semiconductor Jones Act Loom
Published 2026.03.15. 23:31 Updated 2026.03.16. 10:34
Kwon Seok-joon
Kwon graduated with a bachelor’s and master’s from Seoul National University’s Department of Chemical and Biological Engineering and earned a Ph.D. from MIT. He is a professor at Sungkyunkwan University’s Department of Semiconductor Convergence Engineering, researching next-generation semiconductor materials and processes.
Kwon Seok-joon, speaks during an interview with the Chosun Ilbo at the Chosun Ilbo Art Museum in Seoul on the 13th. He said the United States may introduce a “Jones Act for semiconductors” in the future and that South Korea should prepare for it.
On the 12th, the Two Sessions (National People’s Congress and Chinese People’s Political Consultative Conference), China’s largest political event, concluded with the establishment of a new national goal: becoming an “AI (artificial intelligence) economic powerhouse” by 2030. To achieve this, the conclusion was reached that “self-reliance in semiconductor technology” is urgently needed. Days earlier, Chinese semiconductor executives published a policy proposal calling for “mobilizing the entire nation to create a Chinese version of ASML (the Dutch company that manufactures advanced semiconductor lithography equipment).” The aim was to develop domestic alternatives to ASML’s EUV (extreme ultraviolet) lithography equipment, which China cannot import due to U.S. sanctions.
What would happen if South Korea loses its competitive edge in semiconductors? The mere thought is chilling. How advanced are Chinese semiconductor companies’ technological capabilities, and what is their potential for growth? These questions were posed to Professor Kwon Seok-joon of Sungkyunkwan University’s Department of Semiconductor Convergence Engineering. In 2022, Kwon authored “Semiconductor Samgukji,” a book analyzing the global semiconductor hegemony race and survival strategies for South Korea’s semiconductor industry. He is set to publish a follow-up, “China Semiconductor Rising,” in April.
◇China, Systematically Overcoming Technological Challenges
-What is included in your new book?
“It evaluates China’s advanced computing industry strategy and policies, represented by semiconductors and AI, from multidisciplinary perspectives—engineering, business, economics, policy, and international politics. For South Korea, semiconductors and AI are practically the last strongholds. Understanding China’s strategy is critical for the survival and sustainability of South Korea’s advanced industries.”
-How advanced is the Chinese semiconductor industry?
“They have nearly caught up in legacy semiconductor technologies above 10 nanometers. The semiconductor materials, components, equipment, and packaging sectors are also growing rapidly. However, they have yet to reach South Korea’s level in cutting-edge processes. The slogan ‘create a Chinese ASML’ reflects their desire to leap forward technologically.”
-What efforts are being made for semiconductor technological advancement?
“The Chinese government and industry maintain a list of target technologies to overcome, regularly checking progress. Significant advancements have been made. Ten years ago, only 3–4 out of 30 goals were achieved; now, it’s harder to find what “isn’t” working.”
-ASML’s EUV lithography equipment requires over 100,000 parts from 5,000 suppliers. Does China have the capability to produce such equipment?
“Every March, Chinese semiconductor equipment companies showcase new products at ‘Semicon China.’ Last year, I was astonished to see them using domestically developed DUV (deep ultraviolet) lithography equipment, an earlier generation than EUV. A startup even emerged, creating equipment using plasma as a light source (LPP) to bypass ASML’s laser-driven plasma (LDP) method. Currently, China is simultaneously upgrading technologies required for cutting-edge semiconductor manufacturing, such as light source technology, optical equipment manufacturing, and precision mechanical engineering. A Chinese ASML could emerge within a decade.”
-Won’t competing countries respond?
“Chinese engineers say, ‘Time is on our side.’ The generations that built South Korea’s semiconductor dominance—those who entered university in the 1970s, 1980s, and 1990s—are retiring. The 1980s and 1990s cohorts will mostly retire by 2030. If high-skilled engineers and researchers cannot fill this gap, semiconductor manufacturing know-how could be lost. In contrast, China’s semiconductor workforce is very young. Engineers in their 30s and 40s, who have built expertise from the ground up, will reach their prime in 20 years. Hence, Chinese semiconductor scholars say, ‘If we endure, we will win.’”
◇160 Semiconductor Professors, 1,800 Graduate Students at One Engineering College
-How is China cultivating semiconductor talent?
“Take H University in Wuhan: it has over 160 semiconductor professors and 1,800 graduate students. While Sungkyunkwan University, which has a semiconductor contract program, has only one cleanroom for semiconductor fabrication, this university has three. These cleanrooms, filled with hundreds of semiconductor manufacturing devices, require massive power—yet local governments provide electricity almost for free.”
-Does China’s large domestic semiconductor demand positively impact its technological advancement?
“A large domestic market is a significant strength. A decade ago, China’s battery industry was considered inferior to South Korea’s NCM (nickel-cobalt-manganese) batteries, using LFP (lithium iron phosphate) batteries with lower performance. What happened? Through technological development, they overcame drawbacks like low energy density and safety risks, leveraging learning curves and economies of scale to gain price competitiveness, dominating the global battery market. The Chinese model—where the state acts as ‘patient capital’ for 20–30 years, companies accumulate technology through trial and error, engineers gain training, and costs are continuously reduced—might eventually work for semiconductors.”
-Aren’t semiconductors far more technologically complex?
“While we make 10 units, Chinese engineers could test 100 units by operating cleanrooms 24/7 in three shifts, drastically shortening technology update cycles. For China, which has set ‘building an AI economic powerhouse’ as a national task, semiconductor advancement is non-negotiable. Having succeeded in other industries and gained confidence, Chinese leaders believe semiconductors are no exception.”
-Is China’s AI economic strategy related to the U.S.-China hegemony competition?
“Absolutely. DeepSeek, a leading Chinese AI model startup, adopting an ‘open-source’ strategy is telling. While appearing free, using DeepSeek’s models begins a dependency on China’s AI ecosystem. It’s akin to a ‘virtual Belt and Road Initiative,’ where free AI models act as ‘industrial land-grabbing.’”
◇Semiconductor Super Cycle to Last Until Late 2027
-How long will this semiconductor super cycle last?
“I predict until the end of 2027. A key difference from past cycles is memory makers demanding ‘take-or-pay’ contracts. Previously, clients paid deposits but could cancel orders if conditions worsened. Under take-or-pay, once a client orders future quantities, they must either accept delivery or pay compensation. Given Samsung Electronics and SK Hynix’s long-term contracts, the super cycle will likely persist for at least two more years.”
-What should semiconductor companies and governments prepare for?
“A memory foundry-like environment will emerge. For example, HBM supplied to NVIDIA is hard to use directly in Google’s AI semiconductor TPU. Companies must offer ‘memory foundry’ services, providing customized semiconductors tailored to big tech’s proprietary AI chips.”
-Could the SK Hynix-TSMC-NVIDIA triangle alliance break?
“Possibly. In a few years, TSMC might offer to produce not just HBM but also logic dies (core engines for AI chips) that connect memory and GPUs, directly supplying NVIDIA.”
◇China Could Achieve ‘Game-Changing’ Semiconductor Technology
-Could China develop a new concept in memory semiconductors?
“Unable to use NVIDIA’s AI chips, they might create entirely new AI semiconductors. They could attempt risky innovations, like 3D transistor architectures that eliminate the need for EUV, achieving high performance even with processes above 10 nanometers. If realized, this would be a ‘disruptive technology’ shaking the industry and a powerful counterpunch against the U.S.”
-What is needed for the success of the Yongin Semiconductor Cluster, which requires hundreds of trillions of won in investment?
“To avoid idle production capacity, power, water supply, and semiconductor talent must be secured. While automation progresses, the ‘last mile’—final finishing—requires skilled experts.”
-Due to U.S. pressure, Samsung and SK Hynix are building advanced semiconductor factories in the U.S. Isn’t this a risk?
“It’s positive in that Korean companies become key players in the U.S. semiconductor ecosystem. However, South Korea must prepare for potential ‘semiconductor Jones Acts’—regulations requiring semiconductors for major U.S.-led AI projects to be made in U.S. factories. Conversely, Korean companies’ U.S. presence could secure opportunities to participate in future game-changing technologies like quantum computing.”
