Recently, captains of Samsung and SK showed their confidence regardless of the president’s ‘worries on semiconductors’ and revealed that there will be no changes in investment plans. However, issues such as ‘China’s chase’ and ‘reduction in global consumption’ are still under concern since Korea’s export economy heavily relies on semiconductors. Professor Jongho Lee also said, “We don’t have a lot of time left, but we still have our chances.”

At the beginning of this year, there were consecutive news articles saying that the golden era of Korea’s semiconductor industry came to an end. Semiconductor exports declined around 30% earlier this year and Samsung Electronics’ first-quarter revenue dropped by around 7 trillion won. Concerns over the crisis are getting stronger and more are doubting the future, saying that the ‘party’ is over. However, it is also true that judging by last year’s revenue of Samsung Electronics and SK Hynix, the top two companies in memory semiconductors, the business profits are expected to meet 60 trillion won just with semiconductors. The current market status is not depressing, but voices of concerns are raised with the chase of our competitor, China, and the global economic slowdown.

We met Jongho Lee(53), director of SNU Inter-University Semiconductor Research Center (professor of the Department of Electrical and Computer Engineering), at Seoul National University in the afternoon of January 8th. We wanted to ask about what is actually behind the crisis. Director Lee is an IEEE fellow, which is the highest grade of membership given to top 0.1% members, and is the world’s number one authority on semiconductors. He is the one who developed ‘bulk FinFet’ technology, which is a three-dimensional semiconductor component developed for the first time in the world. The technology was acquired by Intel of the United States from which he receives loyalty. It is currently chosen by major semiconductor companies from around the world as the key standard technology.

It is not an exaggeration to say that the path he has walked was in line with how the semiconductor industry of Korea developed. Time has passed and the semiconductor industry has become the foundation of Korea’s economy. According to last year’s data, it took up 20% of the total export revenue which comes to 100 billion dollars. Therefore, the fall of semiconductors will startle not only the economy of Korea but also the entire industry. It is simply not just a problem of earning money or not. We need to perceive the crisis first. Professor Lee focused on the ‘essence’ and explained about the crisis.

Korea’s memory technology viewed in awe by other countries

Director of SNU Inter-University Semiconductor Research Center Jongho Lee said that due to the reduction in research funding in semiconductors since 4~5 years ago, it became harder to hire new professors and renovate research facilities. He emphasized that to be able to hold the standard for technology related to artificial intelligence such as neuromorphic computing based on memory, which we are strongest in, all around support is needed to develop human resources.

Why are concerns on the crisis of semiconductors repeatedly being raised?

It is because Korea has risen as the first mover. Having perseverance and confidence is not an answer anymore. Even with all our possible abilities, it will still not be enough. The executives leading Samsung Electronics and SK Hynix, the current top two companies in the field of memory, were the best engineering graduates of the 1980s in Korea. Back then, top students chose electrical engineering and physics over medicine. Later on, they became top-level researchers and hand-in-hand with executives of major companies who made bold decisions, they developed Korea to be one of the strongest countries in semiconductors.

One of the concerns of the crisis is the ‘heavy reliance in memory’

Korea is strong in the field of memory. Some say that there can be significant consequences if the global consumption of memory declines. It has a point. However, this crisis will not happen in the near future since memory is an essential component in IT processes and the needs in memory devices will only increase from now on. Also, Korea is the only country that tends to differentiate between memory and non-memory. It only has a difference in semiconductor logic (digital circuit) technology. A CPU, which is commonly known as a non-memory semiconductor, is also a special purpose logic semiconductor.

Voices are being raised opting for development in the non-memory area

Saying that we make good memory semiconductors contains more meanings than it seems. It also means that we are good at related materials, components, circuit designs, architecture (system architecture and design), and sales. We have a lot of human resources too. Foreigners are rather envious of Korea. It is all relative.

Voices are being raised opting for development in the non-memory area

We need to contemplate on the direction. Would we be able to establish CPU manufacturing companies like Intel or AMD if we start the research now knowing our weakness in non-memory semiconductors? Maybe if we make use of all possible resources we have in the country. This is also inefficient. Becoming the top in cognitive sensors, non-memories, and artificial intelligence is too much to ask for. We should utilize what we can do well, memories. New applications can open new markets. If memory technology was useless and will disappear soon, China would not be investing so heavily to chase us.

We also often heard of how fierce China is.

It is true. There was nothing to worry about 15 years ago. Universities nor companies had even proper Fabs(semiconductor fabrication plant). However, my thoughts changed every year going and coming from China. The speed of development is too fast. The new Fab I visited at Tsinghua University had fantastic facilities. After I came back, when I was about to say that I am envious about the new cutting-edge facilities, I heard the news that they were building another completely new Fab building. This is not even a year ago.

Is it the same in research and development?

It is more than a surprise. Last year December, the International Electronic Devices Meeting (IEDM) was held in San Francisco, the United States. Since it is a top-level conference in the field of semiconductor technology, it is an honor to have a paper accepted. There were two papers accepted from Korean universities. However, China had multiple times more. Beijing University and Tsinghua University alone had more than 8 papers each. What is more surprising is the fact that the host gave a separate presentation on China in front of around 2000 people. The presentation was about how 35% of the submitted papers were written from mainland China. Not only that, out of the rest 65% percent of the papers, 35% of the research contents were from mainland China. It meant that close to half of the research was done by the Chinese. In the middle of the trade war between China and the United States, this was the official announcement at a top conference of the United States.

Top conference of the United States ‘Half of the papers on semiconductors were written from China’

How is the atmosphere in China?

It can be clearly seen that the government provides tremendous support. I had a chance to talk to one of the student of Tsinghua University, who said that regarding the cut line for each major, electrical and electronics engineering are hardest to get in. It reminded me of the popularity of engineering during the 1980s in Korea. The current leaders of Korea’s semiconductor manufacturers were top engineering students from the 1980s. China is currently in that phase. In the semiconductor industry where the winner and loser are decided through a game in knowledge, this left a strong impression.

There are heated discussions on China’s technology level in semiconductors

It already obtained quite a level, but it will continue to chase Korea faster and faster. It is because they are following our path. When a new Korean product is introduced, they conduct analysis and evaluations in detail. They can speed up development since they have a definite goal. In addition, China keeps multiple options open and invests capital and human resources. This is why China can accomplish what took us 10 years in only 3 years. This is why some people say that Korea accumulated time to develop technology since it is short of resources (capital + human), but China accumulates resources to earn time.

Will Korea be taken over by China eventually?

Both countries have their chances. First, Korea has the world’s best capability of semiconductor manufacturing in the field of memory. It is the result of close to perfect endeavors throughout the decades to catch even the smallest errors. This is hard to achieve with just huge capital and human resources. In the field of memory semiconductors, the top two companies in Korea are 10 years ahead of China in terms of technology. However, this cannot remain assured if Korea does not start investing in the future right now.

China is currently focusing on acquiring a two-digit number in market share. When it achieves over 10%, it can develop its business instead of relying on support from the government. If Chinese companies get more competitive, things can get complicated. On the other hand, if China stays content with the 10% level, it can lose its growth engine and become reliant to the government.

How can we maintain the growth engine of semiconductors?

The government should make a control tower and initiate a ‘New gap’ strategy. Samsung and SK only focused on ‘extreme gap’ strategies that no one else could dream of until now. To increase efficiency, however, rather than simply making an all-out effort, a new market should be created with a new gap strategy which develops applications and slowly increases the performance gap. A big picture should be planned by experts who understand the current structure of Korea’s semiconductor industry and see through the global market status.

Semiconductors is a foundational industry that holds the destiny of a country. One of the reasons why the United States could become the strongest county in the world is because it opened up the computing market by developing the standard of IT technology. Korea already has both big companies that have strong hardware but are in need of new ideas, and academics and small and medium-sized companies that lack foundation but have many creative ideas. If they are adequately connected, Korea can again step up to the world’s top level. There is one more thing. Although China is only considered a competitor now, we need to think about cooperation. Sensitive matters can exist between companies, but co-research and development centered around universities should be initiated. China’s semiconductor development will continue, and if we cannot leave them behind completely, it is also a strategy to hold hands with them. 


What is the most promising area in Korea at the moment?

Research in neuromorphic computing. This technology mimics the human brain to save and process a massive amount of information. It is the chance to develop artificial intelligence technologies based on worlds’ best memory technology. It is even irrelevant to the IT standard that the United States possesses. It is an open market.


The reality is still not satisfactory. Currently Korea is only trying to catch up with Google instead of being the ‘world’s best’. However, professor Yu Wang of Tsinghua University transferred his parallel accelerator technology to Xilinx, a semiconductor company of the United States. In other words, the hardnosed America bought from China a core technology of deep learning software where artificial intelligence is operated.

Is this the reason why Inter-University Semiconductor Research Center was founded?

Yes. We don’t have time to wait for someone to roll up their sleeves. The semiconductor industry needs human resources. That is why a two-week basic course on manufacturing semiconductors has been held as a class of about 140 peoples for university students, graduate students, and industry workers. The participants can touch real semiconductor wafers that cannot be seen anymore due to factory automation. It was incredibly popular. The school server collapsed with thousands of applicants competing for the class at once. There are over 13000 people who completed the course. Because it is a non-profit program, the current education fee is insufficient to even maintain the facilities. The research center clean room is small and manufacturing machines are over 30 years old. Provided budget is not enough so when the time is available, we go out to companies as sales. We tried to convince them countless times for sponsorships as we can provide education for their employees.  


Professor Lee showed his anxiety throughout the interview. He did not emphasize a certain person’s role. He said that we should all move together with the same goal. He said, “We came to a state where Korea’s semiconductor technology should evolve a step further. It is impossible to compete with China for the market man-to-man. The government, academics, and companies of Korea should all come together and move the system with a carefully thought-out strategy.’ It means that this only then can there be a chance of winning. Director of the Inter-University Semiconductor Research Center, Jongho Lee, emphasized the three keywords ‘convergence, knowledge, and creativity’ and said the following.

“The monthly pay as a director is 550000 won. I’m not here to go out on sales to companies to earn money. It is for the future intellectuals that will take care of Korea’s semiconductor industry. This is how all the people in this field including me strived until now. We sacrificed our youth to make the best semiconductors in the world. I am not trying to force such old-fashioned values to current students. A position as the top of the world is both hard to obtain and maintain. It is time for the government, academics, and companies to come together. At the Siege of Ansi, when we fought against Tang’s forces, civilian soldiers also participated. The project to make a control tower, why don’t we call this the ‘Siege of Ansi Project.’

Source: https://ee.snu.ac.kr/community/news?bm=v&bbsidx=48452
Translated by Kyungjin Lee, English Editor of Department of Electrical and Computer Engineering, jin11542@snu.ac.kr