Recently, the US Senate passed the NDAA 2019 ("US Defense Authorization Act") with an overwhelming advantage of 85 votes to 10 votes, which included provisions for ZTE, which will not only prevent the Trump administration from withdrawing the ban on ZTE, the future. It will also regard ZTE and other companies as an extension of China's intelligence organizations, prohibiting the penetration of their products and curbing related technology transactions. Behind a series of aggressive measures is the US government's 2025 development plan for China's manufacturing industry, and the 5G era. The counterbalance and game of the same.
This month, the future forum closed-door training seminar was held in Beijing. Experts from all walks of life expressed opinions on the status quo and future development trend of China's chip industry, talent training and industrial ecology of China's chip industry.
ZTE’s events outweigh the disadvantages
Horizon Robotics founder & CEO, future forum youth director Yu Kai first said that the ZTE incident made the chip become a 'reliable' thing, just like 95% of computer professionals did not do depth 6 years ago. Learning, and because AlphGO detonated and vigorously led the development of deep learning.
Director of the Control Laboratory of the Institute of Computing Technology, Chinese Academy of Sciences, Han Yinhe, a member of the Future Forum Youth Alliance, said: 'After the ZTE event, all of our companies, regardless of their size, value the chip height. This is actually very rare. The chip is The foundation of industry, we have felt that it is very important for a long time, but not many people do it. This time the ZTE incident gave us the opportunity.
Professor of Electronic Engineering Department of Tsinghua University, Wang Yu, a member of the Future Forum Youth Alliance, agreed. He said that we have seen some self-sufficiency in some aspects of the domestic IC industry, but there are still many important industries in the chip. A big gap, such as high-speed optical communication interfaces, large-scale FPGAs, high-speed high-precision ADC/DAC, etc., where there is a gap is where there is a chance.
Domestic chip industry needs improvement
Wang Yu said, what is the status quo of the industry? First of all, from the perspective of people, colleges and universities have not invested enough. So far, at least the industry believes that the level of colleges and universities is not as good as that of the industry. The data in the schools is not enough. The chip industry is the same. Train a person for two to three years. In fact, even if the chip does not have the opportunity to go through the complete process, it will face graduation. Secondly, the continuous innovation investment in the chip industry is not enough. Before the reform and opening up China's chip or semiconductor industry is still quite awkward. After the reform and opening up, distribution, OEM and other ways to make money, have an impact on independent innovation, buried the current pit. Later found that there is already a gap It is difficult to catch up with the reverse design of the chip. Thirdly, the current investment in national research projects has increased, but sometimes the actual results and goals are out of touch. In a sense, this is a waste of money. Under the market mechanism, Capital and customers will test the actual product performance, but in the research project, the inspection is very difficult. The indicator level can be more pragmatic, sometimes the indicator is fixed. Too high.
Rational thinking, giving play to long-term advantages
Yu Kai suggested that in the face of the ZTE incident, we must avoid the nationality of the 'Yihe group'. He said that international first-class is to be realized in the market competition, and must be fully market-oriented. The BAT enterprises with innovative business models are completely different roads. I think that from the first day, our market should be oriented to the international market, because only by fully competing in the global market can a company really grow bigger. Stronger, and the road of nationalization I think will be getting narrower and narrower. In addition, when it comes to the issue of the so-called core key technology's autonomy and controllability, there is no doubt that this is our dream, we hope that China is the key in the entire semiconductor chain. Technically, it can be completely autonomous and controllable. But everyone thinks about it, is this possible? Can it be done within ten years? Today's semiconductors from its core IP, from EDA software, from the entire test system, everything and everything Not in our grasp, this is the first.
Second, the development of high technology, the global division of labor has become a reality, and it is also a trend. Any closed-door car is impossible and harmful. What do we have to do? I think that at our point of concern, We have the opportunity to do something to play our longboard advantage. For example, the 'Horizon' concern - the design of the entire autopilot processor architecture, we hope to make long-board effects, the international division of labor In cooperation, I have the right to speak. I can push my opponent to the negotiating table and talk to me about cooperation. I don’t have any ability to fight back. I think this is a real rational thinking.
Han Yinhe also said that it is very important to make up the short board. It is also important to cultivate the long board as the killer of future competition. In some important applications foreseeable in the future, such as robots, the Internet of Things, etc., when the chips are not mature enough now. By first establishing the ecology and developing the chip system, a 'competition longboard' can be formed.
Knowing the past, breaking the chip dilemma
Researcher of the Institute of Computing Technology, Chinese Academy of Sciences, Director of the Center for Advanced Computer Systems Research, Bao Yungang, the young director of the Future Forum, analyzed the three elements of the US's return to hegemony from the dispute between the US and Japan in the 1970s and 1980s. It is said that if the dilemma of China's semiconductor industry is summarized in three words, it is 'people, money, and things'. The three points of the American solution can be done in China. First, the government should increase investment in basic research and should continue to invest; Second, the supplement of the folk power, such as the future forum is a very good platform, has made a strong supplement; Third, attracting capital, promoting technological transformation, now there are corresponding measures to do. There are two points worthy of attention Concerns: One is talent, attracting overseas talents back; the second is open source chip design.
Bao Yungang said that we have done a statistic. Now there is no shortage of Chinese nationality talents. China has indeed cultivated a lot of outstanding talents. They can do the best work, but these jobs are mainly done in the United States, and graduated. Most of them will stay in the US in the future. So in the future we will attract more talents to return to China. They will have a great effect on the future of China's chip development. Secondly, the design of open source chips, which can lower the threshold of the entire chip design. Today, the threshold of chip design is really very high. It doesn't move a chip design to tens of millions, even hundreds of millions. If our chip design can be as simple as open source software, it is possible to make more large and medium-sized small and medium-sized enterprises. Enterprises also have the ability to design chips, which may overturn the entire chip industry. This is expected, and the United States is already doing it, which is worth thinking about.
Han Yinhe also said that there are serious shortages of talents in the domestic chip field. There are two reasons for this. First, the talent supply is insufficient. The chip talents come from the college microelectronics college. Now the national microelectronics is a secondary discipline, and electronics is a first-level discipline, electronics. There are many secondary disciplines below. All enrollment indicators should be placed in the college's co-ordination. It may not be less than one-third of the total college enrollment indicators, because we have not paid much attention to the microelectronics discipline for a long time. This is the first one. Secondly, the retention rate of talents is not high. Take my students as an example. The graduation direction of chip design is that only one-third or even one-quarter of people go to make chips, and one-third goes to BAT. The other one-third went to state-owned enterprises and other companies. So, the students that we spent four or five years training out of us, most of the last few did not choose chip companies. In recent years, the situation may be better, because there are so many artificial Smart companies, Now my students are a little higher than the proportion of chip companies, but in the past, for a long period of time, few of our students went to chip companies.
In response to these two existing problems, Han Yinhe proposed a four-point response. He said that the first retention rate is determined by the market and it is difficult to change. In the past, the chip industry had low wages. To reverse this situation, only at the national level. The expansion of the supply side can be solved. Second, in terms of talent investment, master's and doctoral training should be focused on, because the chip talent training cycle is very long, and the two-year training model is difficult to cultivate application talents that meet the needs of enterprises. It is found that doctoral students are now more suitable for the actual needs of enterprises, so they can increase the mode of doctoral training in schools. Third, younger and better people are mostly unwilling to make chips. Here is an important issue - evaluation system. The field needs a unique evaluation system that can correctly evaluate the contribution of our field. Fourth, we must cultivate a comprehensive talent that integrates production, education and research. Our national funding model is a national money-making enterprise. There are several types of projects in the US NSF and DARPA. Enterprises are united to set up funds for the project, and the state subsidizes the company according to the situation of the company’s money. This is a very good production, study and research. Mode. Han Yinhe added that there is no such thing as a Chinese computer society in the chip field, or an active organization like the future forum, which is endorsing the chip, resulting in little communication and insufficient activity at all levels of the chip field.
More 'core' sound
Han Yinhe, Director of Control Laboratory, Institute of Computing Technology, Chinese Academy of Sciences
You don't have to be too pessimistic. The chip field is basically the highest peak in the precision manufacturing field in our country. If the chip field breaks through, the industrial system of the whole country will be fine. If the chip problem is solved, we will basically realize the whole industry. The dream of rejuvenation.
Chief Engineer of American Broadcom Company Jin Yi
China is on the rise of development. The ZTE incident is a good opportunity to attract more overseas talents. This is a career development opportunity for them. It is very attractive and is very beneficial to the development of China's chip industry. This is a win-win situation.
Associate Professor, Department of Computer Science and Technology, Peking University
//s3.pfp.sina.net/ea/ad/15/11/0890b346e5eafefeb7bca1f08e13b4cb.jpg In contrast, John L. Hennessy and David A. Patterson, who won the Turing Award in 2017, why do they make chips? They are from their own work. It is found that software and hardware should be coordinated, care about architecture, hardware design, and care for the combination of the two, hoping to attract more people who do OS.
Horizon Robotics Founder & CEO Yu Kai
Share a word, Kennedy said when launching the Apollo program, 'We started the Apollo mission to the moon, not because it is easy, but because it is difficult. 'I feel that crossing this peak, for the entire Chinese technology industry Very important indicator meaning.
Researcher, Institute of Microelectronics, Chinese Academy of Sciences, Deputy Director of Silicon Devices Center
The chip industry is an extreme industry and the crystallization of global wisdom. When the network is up, we have Baidu, Ali, Tencent in the face of Google... In the future, artificial intelligence, the screen will be replaced, 5G network is the center of wisdom, under An architecture will be born.
Future Forum · Closed Farming
Closed-door farming as a high-end closed-door meeting for inter-industry research and capital matching, inviting scientists, entrepreneurs, investors, government and related industry entrepreneurs to participate in the study of cutting-edge scientific research, leading the application of science and technology, leading the future industry Development direction, while participating in the discussion of policy and business model environment, forward-looking, inspiring, speculative, and collision communication.