Xie Zhifeng: China's semiconductor development has the right time, place and people

The article comes from The Paper reporter Chai Zongsheng. Thanks to the original author for his efforts. If the article is not reprinted properly, please contact the backstage in time. Thank you.
A complete semiconductor industry chain is particularly important for China. Xie Zhifeng, CEO of Si Rui Technology, is a veteran in China's semiconductor industry. He joined Intel's first research and development center in 1988. In 2001, he participated in the establishment of Semiconductor Manufacturing International Corporation with Zhang Rujing and worked there for ten years. In September 2012, he founded Shanghai Si Rui Technology Co., Ltd. and continued to work in the semiconductor industry.

Earlier this year, during the filming of the documentary "China Lab", Xie Zhifeng was interviewed by a reporter from The Paper. He believed that the development of China's semiconductor industry currently has the right time, right place, and right people. He explained that the right place means that the market is in China; the right people means that China has enough talents, and both those who have returned from abroad and those who are local have risen. Fifteen years ago, it was necessary to import a decent senior management talent from the United States, and people were unwilling to come. Now such talents are available in China, and we can do well if we seize the opportunity. He said that if we don't seize the opportunity at this stage, we will fall behind in the next cycle. China needs a full industrial chain The Paper: Tsinghua Unigroup is doing a lot of mergers and acquisitions, trying to complete a complete semiconductor industrial chain. So, how important is a complete industrial chain to China? Xie Zhifeng: China needs a full industrial chain, but it is unreliable to put all the sub-industries in one company. Each company has its own genes and its own expertise. It is impossible for a company to do everything and be big and comprehensive to be the world's number one. Only by focusing on one field can it be the world's number one. Every part of the industrial chain needs someone to do it, but it should be separated and different experts should be invited to do it. In history, there is no such thing as a large, comprehensive, and world-class enterprise. Take Intel as an example. My first job was at Intel. Its strongest point is the manufacture of integrated circuit chips. It is still the world's number one today, with no rivals. It is also unique in design, so manufacturing and design are its expertise. But it has no advantage in systems or even mobile phone applications. Its design and manufacturing are biased towards PCs (personal computers), laptops, and cloud computing. This is where its expertise lies. It has made many attempts in its 40-year history. It has shut down and started mobile phone projects, but all have failed. Its genes are suitable for making high-speed and high-performance chips, and it has not been very successful in making low-power mobile phone chips. Maybe it will succeed in the future, but this shows how difficult it is to jump from a field in which it is very good to another field. It is very difficult to defeat others in their field. They are also trying to make achievements in the Internet of Things, but the impression they give to others is the production and design of high-speed and high-performance chips for cloud computing and personal computers. Other than that, others have not regarded it as the best company. Therefore, we also need to cultivate the best company in a certain field, and not expect it to be excellent in all fields.

Pessimistic about mergers of large companies

The Paper: Are you optimistic about the series of mergers and acquisitions launched by Chinese semiconductor companies around the world?

Xie Zhifeng: High-tech mergers and acquisitions are forced by the economic situation. Many companies can no longer survive on their own and must merge together to keep each other warm. Mergers and acquisitions should be complementary. If you have what he doesn't, mergers will be more reasonable. But some mergers and acquisitions are for the sake of mergers and acquisitions, and grouping up does not solve the problem.

I am generally pessimistic about the mergers of large companies. The mergers of large companies around the world have caused talent loss. There are not many successful cases of cutting into small pieces and selling them after the merger, and most of them fail. Enterprises must have a very strong culture. You can integrate other companies after they come in. Sometimes they are nominally a family, but they are full of open and secret struggles. If there is a strong culture, two companies merge together, and everyone agrees, there will be no friction, but most mergers of companies cause conflicts between the merged party and the merged employees, and often the merger is 1+1 less than 2. Only

companies with strong technology and culture can easily integrate others, otherwise the merger will fail. After the cultural shock, a large number of talents are lost. Although some patented technologies can be purchased, the patents are only on paper and need to be implemented by people. It is equivalent to buying a machine, but the people who can use it are no longer there, and it still cannot work. There is no need to distinguish national boundaries for the accessories of terminal products . The Paper: What kind of attraction does the Chinese market have for the companies that make up the entire industrial chain? Xie Zhifeng: The technology market is very cruel. Each terminal product must use the best accessories, and there is no need to distinguish national boundaries. Apple's industrial chain is globally sourced and uses the best components. There is no requirement to use American products. As long as it is self-controlled, there is no problem, unless it is a military product. After acquiring so many companies, the key is whether they can be digested. Companies can be bought, but talents are mobile. Many core things are in people's minds, not in hardware. I have been back to China for 15 years and have seen a trend. China has a huge demand for high-tech products, more than the United States and Europe. 15 years ago, the United States was the largest. Now all high-tech suppliers in the world have come to China to set up factories. Talents, technology, and production are all transferred to mainland China, which has indeed brought about technological migration. Taiwan and South Korea are very envious. They are very small markets, although they are very strong in some areas, such as Samsung in South Korea and TSMC in Taiwan. So they have no choice but to come to the mainland to build factories. TSMC is going to build a world-class factory in Nanjing, Samsung has world-class storage in Xi'an, and Intel is going to increase its investment tenfold in Dalian. They must move to the mainland, otherwise they will lose market opportunities. Mainland companies have the opportunity to develop. These companies will bring talents and technology, and these people will flow in China, whether to state-owned enterprises, foreign enterprises, or private enterprises. In the past, the first priority for graduates from Tsinghua University and Peking University was to go to Europe and the United States, and after graduation, they would work in the best companies in the world, staying at Intel, Apple, and Google. Now many of them stay in China, so there are great opportunities in the future.

Now we need to have confidence to get ahead of the world

The Paper: Are overseas semiconductor companies coming to the mainland because of cost?

Xie Zhifeng: Cost is not the most important thing. Now the cost of making chips in the mainland and in Taiwan is actually about the same. The main thing is the market. You need to be close to customers and you must know what customers want. Far away in the United States, you are not sensitive to the needs of Chinese customers. If you set up a research and development center in China, the response speed will be much faster. So when setting up a factory in China, the first consideration is to be close to customers.

Of course, cost and talent resources are also important. Chinese talents are very hardworking and smart. They need opportunities to develop. To give full play to their intelligence and wisdom, they need good teachers to lead. If there are good companies, whether Chinese or foreign, they will give good training to excellent graduates. Once they have experience, they will become mobile talents in the market.

We must learn the ability to innovate and the methods to solve problems. After learning the methods, we can create new technologies and products. We can have the ambition to create technologies and products that foreign companies do not have. Don't always think about transferring technology from them. The world is flat, and we must build confidence in this regard. In the past, we have always mentioned that we must catch up with and surpass the world's leading level, but most of the time we are catching up, not surpassing, but filling the gaps in the country. Now we need to change our thinking. We should not always think about catching up or surpassing. We should choose the fields where Chinese people have advantages, concentrate our efforts, make breakthroughs, and go ahead of the world. From a strategic perspective, China must develop storage . The Paper: What is the level of segmentation of Chinese semiconductor companies? Xie Zhifeng: The semiconductor industry is divided into design, production, packaging, and testing. In terms of manufacturing level, SMIC is the leader in mainland China, and its level has reached world-class level, and the future is very promising. In terms of design, HiSilicon (HiSilicon Semiconductor Co., Ltd.), headquartered in Shenzhen, has an advantage. In terms of packaging and testing, Changdian (Changdian Technology Co., Ltd.) in Jiangyin, Jiangsu is the largest, and recently acquired Singapore's Jinpeng Technology, reaching a new level, probably reaching the third or fourth level in the world. These three companies can be leaders in China, but what is lacking is the storage industry. In the three key technologies, Intel is the world leader in CPU (central processing unit) technology for logical operations; in terms of storage technology, South Korea's Samsung is the leader; in terms of packaging and testing technology, Taiwan's ASE Group is relatively advanced. There is a big gap between the leading companies in the mainland and them, especially in storage, which is almost non-existent. From a strategic perspective, China must develop storage. The country is preparing to invest 140 billion yuan in five years, but this amount of money is not as much as Samsung and Intel's annual investment combined. The development is very fast in vertical comparison, but compared horizontally, it is still very limited compared with the investment of competitors. And this is not a problem that can be solved by money, but also depends on talent and technology reserves. Don't underestimate patent reserves. As long as you become bigger and gain momentum, you will definitely have to fight a patent war. Once you start fighting, you will be knocked down with one punch. Design is only the design of mobile phone chips. We are still very weak in the design of high-speed CPUs for cloud computing and personal computers. The gap is huge. The gap between mainland semiconductor manufacturing companies and TSMC is very large. In the long run, TSMC will still lead for a long time. This is the result of hard work, hard work, funds, talents, and intellectual property accumulation. After all, it has accumulated for forty years. Manufacturing is difficult. If you want to catch up, you must first solve the problem of money, and then talk about talents, technology, and equipment. Optimistic about the next decade The Paper: The national plan is ambitious. Does China have the conditions to realize this plan? Xie Zhifeng: The current national planning has made progress. It is not just a five-year plan, but also focuses on long-term planning. Made in China 2025 is a ten-year plan. Technology accumulation requires long-term planning. Old friends in the United States are very envious of China's long-term planning. The country has a long-term plan, but people in the industry are eager for quick success and instant benefits. They always want to make a one-shot deal. They want to see results in the second year after investing this year. I am still very optimistic about the next ten years. The right time, right place, and right people are all in place. The right time means that the time has come. The investment in the world is decreasing, while that in China is increasing; the right place means that the market is in China; and the right people mean that China has enough talents, and both those who have returned from abroad and those who are local have risen. Fifteen years ago, to find a decent senior management talent, you must import it, and you must dig it from the United States, but people are not willing to come. Now such talents are available in China, and we can do well if we seize the opportunity now. If we don't seize the opportunity, we will fall behind in the next cycle. The control of national funds is becoming more and more sound

The Paper: There was some discussion about the efficiency of the use of state funds before. How is it now?

Xie Zhifeng: In the past, there was not enough money, and the funds were scattered everywhere. Universities and research institutes all got a share, and there was no synergy. Now we are selecting the best and supporting the strongest. We support the enterprises with strong capabilities. This pattern is much more effective than ten years ago. Now we choose enterprises with hope of success and successful experience to invest. In the past, universities and research institutes were the main ones, but now enterprises are the main ones. Enterprises are leading universities and graduate students, rather than universities and research institutes. This is a big change in recent years. In the past

five years, leaders from the central government to local governments have said that enterprises are the main body of scientific and technological innovation, and research institutes and universities are auxiliary. Because research institutes and universities do not make products, they are out of touch with the market. It is a mature mechanism to decide who to give funds to and who not to give funds to. I dare not say whether it is the best, but now this system is quite sound. It feels that the country is paying more and more attention to it, and the control of funds is becoming more and more sound. State-owned enterprises and private enterprises have the opportunity to get it. Of course, state-owned enterprises have more opportunities, and private enterprises also have opportunities. Especially in our field, whoever has the ability can get it.

The Paper: How to trace the efficiency of investment?

Xie Zhifeng: If your investment has no effect, you will not be given it in the future. If you do not meet the national requirements, your score will be very low when you apply for a project again. This is a very important score. The judges are experts from all over China. They can be leaders of state-owned enterprises or professors of universities. There is an industry expert database, and experts are selected from the database to do projects. This mechanism is quite reasonable. But there are some problems. These experts are all part-time. Sometimes they are too busy and cannot go. Sometimes we cannot find the most suitable expert. If the most suitable one is very busy, other experts have to be used to replace him. However, it is too expensive to support such a group of full-time experts. We pay the experts 500 to 1,000 yuan for travel expenses each time, which is quite in line with China's national conditions. The same situation would be very expensive in the United States. Don't always think about importing, you should innovate on your own

The Paper: What lessons can the semiconductor development experience of South Korea and Taiwan teach the mainland?

Xie Zhifeng: Samsung in South Korea is a huge state-owned enterprise, and many national resources are invested in Samsung and Hynix. In the early days, South Korea bought technology from Japan and sent a large number of students to the United States. After graduation, those students worked in large American companies for five years, and then returned to China to establish Samsung Semiconductor. China does not have such a systematic talent policy. We all go and come back on our own. So the effect of South Korea is very obvious. In the 1980s, I studied in the United States, and those Koreans also studied in the United States. They were all our classmates, and today they are all executives of Samsung and Hynix.

The Paper: The vice president of TSMC was poached. How do you view the secret war in this industry, as well as the exchange of talents and technology?

Xie Zhifeng: The Koreans poached Liang Mengsong, the vice president of TSMC, directly. It really is like a battlefield in the mall, and they fight a business war. The flow of talent and technology, the West has its own rules, but as far as I know, the United States has no explicit ban on selling products to China. It requires transparency and control. You can tell me what technology you want, and they will review it according to their laws and regulations. If they agree, they will give it to you, as long as you don't make weapons to kill people. As long as we maintain good communication, technology is still open. You can buy or learn it through normal channels. Other channels are unwise and have caused many negative effects.

But don't always think about importing. You have people and money, why don't you do it yourself? We need to innovate ourselves, we have the ability. Of course, some things are not available to everyone and need to be imported.

[About reprinting]: Reprinting is limited to reprinting the full text and retaining the title and content of the article in full. No deletion or addition of content is allowed to circumvent the original protection, and the beginning of the article must indicate: Reprinted from the WeChat public account " Semiconductor Industry Observation icbank ". Thank you for your cooperation!

【About submission】: Semiconductor elites are welcome to submit articles. Once accepted, they will be published with your name and rewarded with a red envelope! Please indicate "Submission" in the title of the manuscript email, and indicate your name, phone number, unit and position in the manuscript. Welcome to add my personal WeChat account MooreRen001 or send an email to jyzhang@moore.ren

Click to read the original article to view more high-paying semiconductor jobs