China's total amount of imported semiconductors far exceeds oil, and the road to localization is still a long way to go.
'The integrated circuit industry is the core of the information technology industry and a strategic, fundamental and leading industry that supports economic and social development and guarantees national security.' 'The Outline for Promoting the Development of the National Integrated Circuit Industry (hereinafter referred to as the' Outline ") Outlined the goals and priorities of China in promoting IC industry development.
Globally, as the foundation of the information age, the output value of semiconductors has stagnated growth in the seventh year.While the new breakthroughs in the growth of personal computers and smartphones haunt the industrial giants and entrepreneurs of Silicon Valley in the United States,
For China, since the 1990s, a group of domestic scientists, enterprises and research institutes have been devoted to the research and development of semiconductors. With the support of the "strong core" policy in recent years, including the "Outline", China-made semiconductors have experienced a leap forward Development, but also face the development of technology, capital, equipment, personnel and other multiple problems.
In the view of Chen Datong, the co-founder of Spreadtrum and the chairman of the investment committee of Beijing Qingxin Huachuan Investment Company, the development of China's semiconductor industry has gone through four stages. From the early self-reliance, the government led the gradual development of the government into a model of market operation. The industry The operation of the fund leverages even greater capital power, and the favorable market and policy environment have brought about the industrial status quo of returnees and the accumulation of domestic technical personnel.
Chen Datong was admitted to Tsinghua University on the occasion of his college entrance examination resumed in 1977. He was the first batch of doctoral students majoring in semiconductors in China and later went on to study at the University of Illinois and Stanford University in the United States in 1989. In 1995 and 2001, The United States and China participated in the founding of Omnivision and Spreadtrum Communications, the former dedicated to CMOS image sensor chip, the latter developed a proprietary mobile phone baseband chip, both become the world's industry leader in 2008. In 2008, Spreadtrum listed in the United States, he joined the venture capital, leading a series of investment in the field of integrated circuits.
Since 1978, he has been studying semiconductor specialty and nowadays, Chen Datong has witnessed the growth of China's semiconductor industry and the changes in its position in the world.China's semiconductor industry over the past 40 years of trials and hardships is also the footnote of the 40 years of China's reform and opening up. In recent days, surging news interviewed Chen Datong in Beijing.
'dialogue'
China's semiconductor industry in four stages
Surging News: In 1978, you switched from the Tsinghua University radio major to the semiconductor major. Now that you have been dealing with the semiconductor industry for 40 years now, how do you see China's decades of semiconductor industry development?
Chen Datong: Feel particularly deep.The first stage is before I go abroad, all domestic self-reliance.
China's semiconductor industry began in the late 1950s and early 60s of the last century. At that time, the gap between China and the rest of the world was not large and started almost simultaneously. However, due to two things, one was the 'Cultural Revolution' and the other was 'closed-door state' All the technology, technology, equipment are all done by themselves, slowly opened the gap by the international advanced level.
By the early 1980s, when I was a graduate student, China was already far behind the international standard, when two research institutes, the Institute of Microelectronics of the Chinese Academy of Sciences and the Institute of Microelectronics of Tsinghua University, were the top researchers in China, but Tsinghua University University Institute of Microelectronics to do only one thing, called 'reverse design', in fact, is the anatomical chip, analysis and then re-made the same as the year Intel CPU8088, 8086, hundreds of thousands, millions of transistors, I can To analyze it and to do it, this ability is remarkable, but it does not have any positive design capabilities. This reverse design is to 'catch up' and to break the embargo abroad, but it is not true Super '.
The second phase began in the late 1980s and ended in the 1990s, with the government trying to introduce foreign technologies through the will of the state. There were two national projects at that time, one being the '908 Project' and the introduction of the six-inch production line and the other being the '909 Project' , Introducing the most advanced eight-inch production line of the time.The 909 project was almost the largest project in the country at that time, with an investment of about a hundred billion and wanted to introduce state-of-the-art equipment to develop its own industry, but in the end it proved to be not a successful model because It is also a completely state-owned, with no incentive mechanism and requires constant blood transfusion from the state. The 908 and 909 projects basically proved that doing so in a planned economy is not feasible and there is no market competitiveness.
The next phase starts in 2000 and runs until 2013, marking a few things at this stage.
Two of the most important things took place in 2000 when the foundry of SMIC, which is truly a joint venture with the world, was established by SMIC and Grace, so that the products of chip design companies could be processed domestically. At the same time, At that time, Ministry of Information Industry issued Document No. 18 ("Encouraging the development of the software industry and the integrated circuit industry, a number of policies.") In fact, the document No. 18 is to play a major role in the call, it allows everyone to see the country's attention and support, Many returnees are to see the 18th document, coupled with the foundry in the country, and then come back to do design companies.
There is one more thing, starting from the beginning of this century, especially after 2005, Silicon Valley VC (venture capital investment) began to invest in. With funds, foundries, with national policy support, these three to enable students to Determined to come back and start business. Spreadtrum is like this.
This period of time the semiconductor industry has mushroomed out of many companies, mainly private enterprises, are basically supported by the VC, together with some government support, but government support is not dominant. Although these companies start relatively low, but basically Is to compete directly with large foreign companies because of their efficiency, access to gas and technology, so they have developed very fast. All successful companies have defeated foreign companies in various fields and made use of import substitution models. Before 2013, there were some spontaneous enterprises in the design and packaging, including equipment and materials, and in various industries, which meant that there was a spontaneous industrial foundation.
Surging News: What happens after 2013?
Chen Datong: The more we do, the more we find a problem. Among the entire semiconductor industry, some are non-governmental capitalists, local governments, or can not do it in a completely market-based way. We have to rely on state support to guide development. This is actually The strategic needs of the development of national information industry.
There are several major aspects. For example, memory, investment is very large, slowly leaving only the few giants in the world, the strength is very strong, we rely on private capital support is almost impossible to start from scratch. Must have five to ten The investment period of this year certainly will not make any money. After that, it is not always possible to make any money, but this is especially required by the industry. Therefore, this can only be done by the state in coordinating resources and organizing research.
Another example is the new foundry.Construction is a high investment but the return is very slow thing, in fact, although SMIC made, but the middle is very painful.If you want to reinvest a SMIC, I am afraid that none of private capital Willing to vote, because its return is too slow, the process is painful.
Around the time of 2012, four people wrote a letter to the Central Government, including Li Lianyu, CEO of Spreadtrum, Zhang Wenyi, then chairman of SMIC, Yin Zhiyao, the founder of Semiconductor Microelectronics, and the then Shanghai Semiconductor Association The Secretary-General Jiang Shoulei. This letter has been reported to Shanghai through the top level. After being instructed, the government decided to make the integrated circuit as a national strategy to do, so in 2014 released the "Outline."
This is a very big thing, with two of the most substantive measures. The first is to set up a leading group under the leadership of the Vice Premier of the State Council, Mackay. There is no higher level in developing an industry in all aspects The second is the establishment of a national integrated circuit industry fund (large fund), the Ministry of Finance out of about 1/4 of the money, the other by the CDB and State-owned enterprises pay, operating mode is guided by the state plus the operation of the market to equity investment to do.
Since then, that is, from 2013 to 2014, it will be a new stage, and at this stage, as soon as the country launches, many things that China did not dare to do and can not do can all be done and the development of semiconductors becomes a hot spot. In fact, It is a bit overheated, but it is always a good thing to be indifferent to the previous situation, and it is with this that we can do something that we would not have imagined.
'Entrepreneurs shift from the sea to domestic talent'
Surging news: many places in the past two years are the main push IC industry, does indeed appear as you said signs of overheating. How will this industry heat development?
Chen Datong: In China, any industry will have several stages, the earliest is not, a poor second, suddenly one played an exemplary role, hula look, followed by all-up, and immediately overheated.
However, the planned economy is different from the market economy. The planned economy depends on the plan. There is neither overheating nor development. Under a market economy, everyone will go over the road after overheating. Finally, some particularly viable and truly good enterprises remain. With the market out of the way, and then through the restructuring of mergers and acquisitions, the industry can gradually mature, and the rest are very leading enterprises with strong vitality.
Over the past few years, there have actually been two overheating phenomena in the pan-semiconductor industry, the solar energy industry, which was particularly hot in 2008 and 2009 when China's polysilicon production capacity was three times the world demand, and then the price was shattered , A lot of business losses in the shuffling process, those European and American enterprises almost all hit.Another is the LED industry.In 2016 around the time, a local government suddenly launched a policy, companies buy a MOCVD equipment, a government subsidy Afterwards many places all the same policy, then appeared in the hundreds to do the LED chip company.In the second year, because not so big the market demand, the equipment start-up rate is less than 30% This area is the world of Chinese companies.
First, the semiconductor is a star among the stars, high-tech high-tech which is the most important part of the country's strategic emerging industries in the past who always support the semiconductor industry, the early development of the semiconductor industry, local leaders Almost all of them are reused, and now the country places special emphasis on innovation and the overall climate is greatly improved.
Another difference is that in recent years there has been a fundamental change in the national investment system, with almost no state-to-state investment by 2013, no private-public investment, no overlap between state capital and social capital, and investment in state capital The efficiency is very low during the global financial crisis in 2009, the national 4 trillion investment is one example.
Since 2013, the government has tried a combination of country-led and market-based operations, with the National IC Industry Fund as the first attempt to operate in this way at the national level. To the local government level, many local governments have also been established Semiconductor or IC industry funds, rarely by the state, by the local government to invest wholly-owned projects.
Now that a dollar can be levied by the state at least three bucks and four bucks, and that investments all have market-based assessment indicators rather than a rash decision, this is a fundamental reform. In the past rent-seeking space was too large. Now The right to approve the fund level, the decision by the professional fund managers, the gray area is almost gone.
At the beginning of the founding of the Great Fund, in fact, the Central Government also wanted to command a nationwide chess operation. At that time, it designated "Three plus One" cities (Beijing, Shanghai and Wuhan plus Wuhan) in preparation for the vigorous development of the integrated circuit industry in these places and avoiding duplication of investment. Competition for resources.However, the local development of integrated circuit industry enthusiasm is very high, there has been a flourishing situation.Later to understand that this is a good thing, and finally can be resolved through the way of mergers and acquisitions, investment will not be wasted, so the correct way to guide, Instead of using plans.
Surging news: Earlier years in the country began to engage in the semiconductor industry, many of them are like you returnees. Entrepreneurship in recent years have changed?
Chen Datong: Ten years ago, about 80% of the semiconductor companies are returnees venture.
It turned out that the domestic semiconductor industry is one hundred percent state-owned, the last century 90's foreign semiconductor companies to enter the Chinese market, most state-owned enterprises have been washed away, leaving only some of the military enterprises.After more than a decade the development of the semiconductor industry, the basic private enterprises At that time returnees have the advantage, he came back with technology and products, mainly technology.On the other hand, due to overseas companies done in large foreign companies, VC is more easily recognized.This is why, a decade ago to returnees venture Mainly.
However, returnees also have a natural disadvantage, that is, they do not understand the market, especially the domestic market, so they generally have to go back a few years after returning home. They like to use some of the world's leading technologies to make the world's leading products, China does not have the world's leading customers, so their products can only target customers in Europe and the United States, but small business startups in Europe and the United States basically play, many companies died at this stage.A part of the returnees can want to understand that domestic customers need What, using high technology to make a product adapted to the domestic market, so the company developed.
After these ten years of development, domestic engineers have been trained, but also have seen entrepreneurship is one thing, and now is the domestic talent-based business.The biggest advantage is that they know the local market is very good. In addition In recent years, not all VCs have returned from overseas. Many domestic funds have also contributed to the current situation of domestic entrepreneurs.
'The decline of Silicon Valley is worth learning'
Surging news: When you founded Howell Technology in Silicon Valley in 1995, Lenovo just emerged as the 'Niu Ni dispute' in the country. One insisted on taking the commercial route and the second took the technical route. At last, Lenovo chose the commercial route. What do you think of the choice?
Chen Datong: In the past 30 years, domestic private enterprises have basically developed along the road of 'trade and technology' and need to go up one step. The first generation of companies are mainly trade-oriented, with Lenovo and Huawei all from Lenovo is the second generation of Chinese companies, mainly manufacturing and processing.However, regrettably, Lenovo has not been able to 'technology' this is the problem.This is its problem.Why Huawei did so well, because Huawei last Is a technology-driven, third generation company.Currently, Huawei is developing to the fourth generation of companies (innovative companies), Huawei cell phone has emerged some innovative features.
Of course, Lenovo was not wrong at the time.A large number of domestic successful companies such as Haier, Hisense, Changhong and Konka are almost all such companies.These companies have the characteristic that the technology research and development ability is relatively weak, but the production capacity is particularly strong, Chinese companies finally know what is quality assurance, what is cost control, what is the service to customers, and none of these concepts were originally used. But you just imitate and make a product that has no characteristics and can only rely on price wars.
An important indicator of price war is gross profit. Maori is a fundamental indicator that distinguishes second-generation companies from third-generation companies. Anyone with a gross profit of less than 10% belongs to a manufacturing company with net profit at 1% and 2% Gross margins are so low that we can not afford to do research and development and create a vicious cycle.
However, at that time, Lenovo needed to directly focus on technology research and development, and they had already died on two steps in advance. Because technology research and development are regular, it is hard to make technology research and development, and often it costs a lot of money to burn or die, The risk is relatively small to do, anyway, you can sell done.Any company you go to see how much it is R & D investment, you can determine whether it is the technology to win or manufacturing to win.
The third generation of companies rely on technology research and development to win, the fourth generation of companies rely on creativity to win.For example, like Apple, Apple is characterized by what is done on the original market, but its products come out completely with others Like the subversion of this industry, the most difficult to win by creativity, the semiconductor industry has not reached this stage, but sooner or later there will be.
Surging news: China's semiconductor industry mainly rely on import substitution, when is it possible for us to get out of this situation, become themselves to develop a new product, rather than learn from others, imitate others?
Chen Datong: rely on or imitation is just a transitional stage, but really want to survive, to develop well, we must have its own innovation. Innovation is generally a small, suitable for local functions of the micro-innovation began .For example, Spreadtrum, Spreadtrum's An innovation is dual-card dual standby.China's dual-card dual standby needs appeared on the function machine, so Spreadtrum's single-chip dual-card dual standby on the fire up.
At the beginning of imitation, just like learning to draw, first copy copy, but to a certain stage must be innovative, who innovate who can go further .Now, many domestic semiconductor companies have already begun with innovative products This is a good trend.
Surging News: Does Silicon Valley have any experience we can look at or reference? The semiconductor industry in South Korea and Taiwan of China has rapidly risen in these years. Do they have any experience that we can learn from?
Chen Datong: Silicon Valley in the United States is worth our study, but its semiconductor industry is basically a dozen years ago, from six years of last century to nine years of the last century is its glorious period, but after 2000 It is decaying.
The biggest reason for the decline in the semiconductor industry in Silicon Valley is that the manufacturing industries such as television, refrigerators and mobile phones all moved to Asia and moved to China because of the weakness of the U.S. industry. What should the company do if its customers moved away? Returning to their hometowns, some have also succeeded because the market came to China, which is one of the things we should learn from.
Over the past few years, mergers and acquisitions of major European and American companies have shown a clear trend. Dozens of big companies ended up with only a few remaining ones. The model in Silicon Valley has gone on indefinitely. It has to rely on mergers to counter China. Before 2013, the semiconductor industry There are very few international mergers and acquisitions that have taken place in recent years due to China's stimulus after 2013. If Qualcomm reunites with Broadcom, the remaining fabless, design firm It is very few, leaving only IDM (Design and Manufacturing Company), so the process of decay in Silicon Valley is now, and China is still very prosperous.
Another very useful reference is Japan. Japan was particularly powerful 20 years ago, and now its semiconductor industry almost completely annihilated, for the simple reason .20 years ago, Japan's electronics products all over the world, Sony, Panasonic and other five brands rule Global, so its chip manufacturers supply their own electronics manufacturers on the line.Now most of those electronic products are South Korea and China Taiwan, mainland China tripartite hit, customers gone, resulting in Japan's entire semiconductor industry annihilated.
Short term I think the most powerful is South Korea, South Korea's development model is that we can learn from. In order to do memory, South Korea's Samsung in front of about 10 years has been a loss, behind the fact that a lot of support from the country, after doing so, Samsung invincible , Defeated European and American enterprises, defeated Japanese companies and monopolized the world.