K-Semiconductors Face a Five-Year Test of Technological Edge

"Expanded U.S. HBM Controls Aim to Check Chinese AI"

Korean Firms Must Leverage AI-Driven Demand to Fend Off Chinese Pursuit

Editor's NoteThree years after the publication of "Chip War," the semiconductor industry now stands at the forefront of global economic order and national security supremacy. The extent to which a nation can develop and secure core chips is becoming the determining factor in its geopolitical success. The next chip war is expected to center on "even fiercer competition with Chinese companies" and "talent acquisition." Chris Miller, Tufts University professor and author of "Chip War," stated in an interview with The Asia Business Daily that responding to the rapid rise of China should be the highest priority. Dan Hutcheson, Vice Chairman of TechInsights, predicted that as the thirst for talent capable of designing, manufacturing, packaging semiconductors, and building fabs and data centers intensifies, competition will escalate even further. In a situation where the United States, China, Taiwan, and Korea are all striving for semiconductor supply chain independence, the shortage of skilled talent is bound to become a new bottleneck in the race for AI semiconductor supremacy. SK hynix's recent decision to drop academic requirements and ramp up rolling recruitment is also closely related to these industry shifts.

"The biggest challenge facing Korean companies is the pursuit from Chinese rivals such as ChangXin Memory Technologies (CXMT) and Yangtze Memory Technologies (YMTC)."

Chris Miller, the American professor at Tufts University and author of the best-selling book "Chip War," stated in a written interview with The Asia Business Daily that the greatest challenge for South Korea's semiconductor industry in the AI era is maintaining its technological edge over Chinese memory companies. While China is rapidly expanding its market share based on national subsidies and a vast domestic market, Miller described Korea as standing at a crossroads—needing to maintain its technological advantage by leveraging the soaring demand for advanced memory driven by AI proliferation.


Chris Miller, American professor and author of "Chip War"

Chris Miller, American professor and author of "Chip War"

View original image

Professor Miller said, "The United States, China, and Taiwan are all moving to reduce their dependence on global supply chains, making competition even fiercer," adding, "The next five years will be a critical period during which the technological edge of Korean semiconductor companies will be put to the test."


He particularly pointed to the rapid growth of Chinese memory companies as the most direct threat. According to him, "The biggest challenge facing Korean companies is the pursuit from Chinese competitors such as CXMT and YMTC. Powered by massive state subsidies, these companies are expanding their market share both inside and outside of China."


Professor Miller also highlighted the risk posed by Chinese firms gaining price competitiveness through government support. He explained, "As in other industries, Chinese companies, thanks to state backing, can sell their products at significantly lower prices than their overseas competitors."


However, he pointed out that the biggest problem China still faces is access to state-of-the-art semiconductor manufacturing capabilities. Despite rapidly growing its memory industry backed by state subsidies and its domestic market, China continues to face limitations when it comes to cutting-edge manufacturing capabilities.


According to Professor Miller, the key reason Korean companies have maintained their current competitiveness is AI. He said, "Samsung Electronics and SK hynix owe much of their current competitive edge essentially to AI."


The proliferation of AI has also reshaped the strategic value of memory. Miller said, "When 'Chip War' was first published, logic chips represented the most obvious bottleneck in advanced computing. But AI has demonstrated that computation alone is not enough; memory has also become a crucial component."


He continued, "Today's AI systems require the ability to store and transfer massive amounts of data. As a result, AI has greatly increased the demand for high bandwidth memory (HBM)." He further evaluated, "Samsung Electronics and especially SK hynix, both leading producers of advanced memory, have seen their strategic significance grow substantially and are now leading the market."


With the increased strategic importance of memory, Miller analyzed that U.S. regulations on semiconductors targeting China have expanded beyond logic chips to include HBM. He noted, "Policy authorities initially focused on logic chips, which were the most prominent bottleneck in AI. However, as AI models have started requiring greater memory bandwidth, authorities recognized that restricting only processors is not enough to fully limit AI capabilities."


However, he also assessed that China gained valuable time during the delay in implementing HBM regulations. "During the gap in export controls, China managed to stockpile HBM and developed software to integrate HBM more effectively into Chinese systems," he commented.


The competition over AI semiconductors is now expanding from the industrial sector to the realm of national security. Miller emphasized, "The rise of AI has further strengthened the logic I presented in 'Chip War.' Access to state-of-the-art semiconductors is now becoming a prerequisite for next-generation military capabilities."



He explained, "The efficiency of the modern battlefield hinges on the ability to rapidly process enormous volumes of data, enabling efficient decision-making and supporting autonomous systems. What's changed is that semiconductors now form the foundation for AI models that power these systems." He added, "As a result, controlling advanced semiconductors and memory has become more important than ever before."


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing