[Interview] Shin Hyuncheol, Next President of Semiconductor Engineering Society: "2040 Roadmap is a Guide to Competitiveness"

Establishment of Vision Strategy Committee within the Society... Roadmap to be Developed over One Year

"The government has also established a semiconductor technology roadmap, but there are limitations due to the need to handle various tasks such as short-term policy formulation and research fund allocation. The idea started from the belief that the academic society should play a role in discussing and presenting long-term technology development and workforce training roadmaps from a neutral standpoint."

The Semiconductor Society of Korea, the only domestic academic society specializing in semiconductors, recently gained attention by announcing a 'Semiconductor Technology Roadmap' that outlines the outlook for South Korea's semiconductor technology 15 years from now. This is the first mid- to long-term semiconductor technology roadmap in Korea.

Professor Shin Hyun-chul, head of the Department of Semiconductor Systems Engineering at Kwangwoon University and newly elected president of the Semiconductor Engineering Society, is being interviewed by Asia Economy in the Dean's Office of the College of Electronics and Information Engineering at Kwangwoon University in Nowon-gu, Seoul. Photo by Heo Young-han

The person who led the roadmap creation is Professor Shin Hyun-chul of the Semiconductor Systems Engineering Department at Kwangwoon University, who will serve as the next president of the Semiconductor Society. Professor Shin, along with seven colleagues including fellow professors and Samsung Electronics researchers, established a Vision Strategy Committee within the society in January this year and led the roadmap development for one year. In a recent interview with Asia Economy, he explained the motivation behind the roadmap, saying, "It is intended to serve as a guide to maintain global competitiveness by setting a standard of 'this much capability is necessary.'"

The roadmap is based on the premise that by 2040, semiconductor process technology must reach the 0.3 nm (nanometer, one-billionth of a meter) level to maintain competitiveness in the global market. It includes development directions for ▲ semiconductor devices and processes ▲ artificial intelligence (AI) semiconductors ▲ optical interconnect semiconductors ▲ wireless interconnect semiconductor technologies. For AI semiconductors, it is expected that by 2040, learning tasks will be able to process 1,000 trillion operations per second and inference tasks 100 trillion operations per second with just 1 watt of power?about 100 times improvement compared to 2025.

Professor Shin shared the response after the roadmap announcement, stating, "I received many contacts from industry and government officials saying, 'Let's discuss parts we can collaborate on,' and I also received a lot of encouragement that 'this is truly the role the academic society should play.'"

Semiconductor roadmaps are already actively developed in major countries. The United States began semiconductor technology roadmap activities in 1992 when the Semiconductor Industry Association (SIA) established the National Technology Roadmap for Semiconductors (NTRS). This evolved into the International Technology Roadmap for Semiconductors (ITRS) in 1998 and the International Roadmap for Devices and Systems (IRDS) in 2016. The IRDS includes participation from the European SiNANO Institute and Japan's System Device Roadmap Committee (SDRJ). The Korean government has also established a 'Semiconductor Technology Roadmap' through the Ministry of Science and ICT, aiming to secure core technologies for the next 10 years, announcing 45 key technologies in device, design, and process fields.

Professor Shin plans to release an additional formal roadmap of about 100 pages later this month. He stated that the roadmap will undergo major updates every three years, with detailed fields added annually, expanding the key technologies from the current four to ten.

Professor Shin, a former Samsung Electronics researcher, also shared his views on current semiconductor industry issues. Regarding the exemption of the 52-hour workweek rule for semiconductor R&D, he said, "Considering national competitiveness, work hour restrictions should be limited to production areas," adding, "It is unrealistic to impose work hour limits on R&D personnel." He explained, "Due to the nature of R&D, the approach of 'working only this much and then definitely resting' is inefficient." He also mentioned, "An executive who participated in HBM development at SK Hynix said that during the HBM development, he did not leave the company building for 3 to 4 months."

On memory technology trends, he explained, "The market is shifting to a small-scale, application-specialized multi-product market, and the concept of on-device AI is emerging." He added, "Unlike Nvidia's GPU, on-device AI is a small, lightweight system rather than a large system. If this market expands, HBM may not be necessary." This suggests that just as GPUs and HBM, which optimize their performance, suddenly rose due to increased AI demand, another memory technology could replace HBM.

Profile

▲ 1987?1998 KAIST Bachelor’s, Master’s, and Ph.D. ▲ 1997 Researcher at Daimler-Benz Research Center ▲ 1998?2000 Senior Researcher at Samsung Electronics System LSI Division ▲ 2000?2002 Postdoctoral Researcher at UCLA, USA ▲ 2002?2003 Senior Researcher at Qualcomm ▲ 2003?Present Professor at Kwangwoon University Semiconductor Systems Engineering Department ▲ 2010?2011 Visiting Professor at Qualcomm ▲ 2010?Present Senior Member of the Institute of Electrical and Electronics Engineers (IEEE) ▲ 2023?Present Dean of the College of Electronics and Information Engineering at Kwangwoon University ▲ 2024 Senior Vice President of the Semiconductor Society ▲ 2025 President of the Semiconductor Society (designated)