◆ Seung-Woo Seo, the director of the Korean Institute of Electronics and Information Engineers and a professor at SNU

There is a chronic shortage of manpower across strategic industries such as the semiconductor and the battery industry

Current admission with majors and tinkering with talent development will not be able to satisfy future demands

There is a need to break down barriers to expand the talent pool and create an ecosystem

There is a dire need to secure super-gap technologies… The government needs to act as a bridge between industrial and the academic sectors

In the first quarter of this year, the share of Chinese companies such as SMIC in the global semiconductor foundry market was 10.2%, exceeding 10% for the first time. Despite the US sanctions, China is accelerating its semiconductor development and is fiercely catching up to Korea. Taiwan’s TSMC is leading the market and widening the market share gap with Samsung Electronics. Following the warnings that ‘Semiconductor Korea’ is in crisis, the government is coming up with countermeasures such as manpower development policies and regulatory reforms. In an interview with Seoul Economic Daily on the 11th, Seung-Woo Seo the director of the Korean Institute of Electronics and Information Engineers and a professor at the SNU Department of Electrical and Computer Engineering, said, “The government’s talent development program will not be able to satisfy the future demands”, and that “through policies such as major-less admissions, there is a need for university education reforms to expand the talent pool.” He also warned that “China is backed with proper government support, manpower, and ecosystems, and thus could catch up to Korea within 5 years.”

-What must be done to develop the competitiveness of the semiconductor industry?

△ There are three main points. The semiconductor industry is a device-based industry that requires a lot of investment, but it is also a key industry, so government support is more important than anything else. However, financial support is not enough for the industry’s prosperity. It needs personnel to be run properly. Ultimately, it is a matter of manpower, which includes professors and students. Lastly, there is a need for a solid infrastructure, that is, an ecosystem.

-What is the situation in Korea?

△ Several problems emerged over the last 10 years. First is the government’s attitude. The semiconductor business is for large corporations such as Samsung Electronics and SK Hynix, and there was a perception that the government does not need to pay much attention to the industry when those companies are already doing well. The government thus neglected support for the industry to some extent. The same goes for the manpower aspect. For reasons such as overcrowding in the capital area, the expansion of manpower in certain areas such as the semiconductor industry was not active. The creation of a proper ecosystem was also left entirely to large companies; the government did not make an effort to create one. Looking at these three aspects, it can be said that the Korean semiconductor industry has been stagnant for the last 10 years.

-In the meantime, China has been catching up to Korea at a fearsomely fast pace.

△ China is better than Korea in three aspects: government support, manpower, and the establishment of an ecosystem. Currently, there are about 2,000 to 3,000 Chinese companies designing AI-related chips. Most of these companies are funded by the investment of money that they received from the Chinese government. In contrast, Korea only has 5 to 10 such companies. In short, a comparison cannot even be made. The starting point is different. Under these circumstances, it is only a matter of time before China catches up to us. I believe that it will take China about 5 years, and have to come up with countermeasures during this period.

-Why did our fabless field fall behind so much?

△ The fabless field is mainly driven by universities. Universities do not have fabrication equipment as companies do, so they focus on disseminating theoretical knowledge. Students in turn design various semiconductor chips based on the learned theory. Foundry companies play the role of manufacturing the designed semiconductors. Through this process, feedback on the design is provided, and the design competency of the students can be improved. This activates the fabless field, and a desirable ecosystem is formed. But the conditions in Korea are poor. When design samples made by students are sent to a domestic foundry company, it takes more than six months for the chips to be manufactured, since designs made by universities are pushed out of the priority queue. This is largely due to the lack of production capacity of domestic companies and a weak industry-university cooperation mindset. The reality is that many domestic professors in the semiconductor field are commissioning foreign companies such as Taiwan’s TSMC to produce samples. In order to quickly educate semiconductor personnel, it is necessary to design, test, and improve on self-designed chips, but Korea does not have the proper environment to facilitate such activities.

-What is the situation in other countries?

△ If you send a design sample to TSMC, you will receive the fabricated product in about 6 weeks. TSMC receives orders from all over the world, even from China, and produces the commissioned product. Consignment production is a business where one does not know when, where, or what kind of orders will be requested. However, if a foundry company claims to be able to survive on receiving orders only from large customers, and only handles orders from them, universities and SMEs will be forced out of the priority list. TSMC may have a priority list, but it has sufficient production capabilities to handle requests from universities and is also active in industry-university cooperation. Here lies the reason why the Taiwanese semiconductor industry is strong. There is a virtuous cycle taking place within the semiconductor ecosystem that TSMC has constructed. China is also well-equipped from an ecosystem point of view.

-What are the fundamental solutions to facilitate manpower development?

△ There needs to be a reform in the university education system. It is necessary to promote a multi-disciplinary education by introducing a major-less admission system, though it may seem slightly extreme. Rather than accepting students into divided majors prior to educating them, students should be able to learn basic courses related to several majors provided by each college until the first or the second year. There is a need to change the system to one where students will receive their bachelor’s degrees when they enter the 3^rd or the 4^th year with a specified major and meet the graduation requirements. In this way, students’ understanding of various technologies, and their ability to converge across multiple disciplines will be maximized. The strategy for nurturing industrial technological manpower must also be strengthened in the direction of expanding the pool of potential technological manpower. In the industrial field, there is a need for an abundant supply of manpower that can quickly acquire new technologies and develop new products and services when needed. When the pool of potential technological experts expands, companies will be able to develop personnel to their needs at a low-cost through on-site training sessions. To this end, the government needs to provide a permanent systematic and comprehensive career change ladder program so that those that majored in other areas can also contribute to the industry by entering the required fields. Universities and affiliated research institutes should be actively utilized.

-Securing super-gap technologies is a hot topic.

△ Recently, Samsung Electronics announced that they will apply the 3-nano technology to their fabrication process. Such technologies are super-gap technologies, which are essential to outperform competitors such as China and Taiwan. The importance of acquiring such super-gap technologies cannot be overemphasized. However, though semiconductors do require advanced technologies such as the 3-nano fabrication, the semiconductor industry still does require many technologies developed in the past. Automobiles can be built with 10-nano fabrication processes without problems. Along with a strategy to secure a technological super-gap, it is also necessary to establish a market maintenance strategy that would allow Korea to lead the market.

-To develop such super-gap technologies, it seems that cooperation from the government, industries, and universities would all be required.

△ To develop skilled labor in the field of cutting-edge technology, the government needs to step in to create an ecosystem. In the case of system semiconductors, the number of design companies must increase, since the industry is based on a system for small volumes of many products. It is necessary to create an infrastructure that allows fabless companies to freely design prototypes and mass-produce products in a short period of time. If such an ecosystem were to be created, manpower will be trained at a much higher rate. The government has to actively provide financial and tax support to secure human resources and train skilled labor, and act as a bridge that connects industries and universities.

◆He is…

Born in Daegu in 1964, he graduated from the SNU Department of Electrical Engineering and received his Ph.D. from Pennsylvania State University in Electrical Engineering. After working as an assistant professor at Pennsylvania State University and a researcher at Princeton University, he served as a professor at the SNU Department of Electrical and Computer Engineering since 1996. Since 2009, he served as the director of SNU’s Vehicle Intelligence Laboratory and took office as the director of the Korean Institute of Electronics and Information Engineers in January this year. His major publications include ‘The Engineer’s Time Travel’, ‘Security Economics’, and ‘Time of Accumulation’(co-authored).

Source: https://ece.snu.ac.kr/community/news?bm=v&bbsidx=52786

Translated by: Do-Hyung Kim, English Editor of the Department of Electrical and Computer Engineering, kimdohyung@snu.ac.kr