On June 7, 2018, the U.S. Department of Commerce officially announced a new settlement agreement with ZTE. The United States will withdraw the ban on ZTE. On the 19th, according to CNBC, the U.S. Senate passed the "National Defense Authorization Act," which includes a ZTE will continue to implement the provisions of the sanctions. However, this bill still needs to consult with the Chamber of Deputies. It is not clear whether the ZTE sanctions will be in the final version.
Lei Fengnet previously reported that although the previous settlement agreement was reached, ZTE will pay a terrible price: 30 days to replace the board of directors and management; to pay a fine of 1 billion US dollars; to provide a deposit of 400 million US dollars; to accept the US-selected assistant compliance Official, for a period of 10 years.
The ZTE incident caused a sensation in the country, media, the public, experts, and all parties in the industry to join in the discussion of this incident. One view is that although the cost is huge, the ZTE crisis has finally been solved, saving a company with a scale of 100 billion. There are also people. It is believed that the US Department of Commerce’s agreement is the commercial version of the “behaviour of power” clause, and ZTE should be directly disbanded and reorganized.
This incident also triggered public concerns and discussions on the plight of the semiconductor industry. On the morning of the 16th, the China Computer Society’s CCF YOCSEF organized the “China Semiconductor Industry Dilemma and Solutions” forum, inviting industry insiders and experts and scholars to surround the ZTE event and semiconductors. The issue of industrial development issues opinions, suggestions and suggestions, with a view to proposing feasible solutions and follow-up measures to solve the industry's difficulties. Lei Fengnet also came to the forum.
ZTE Incident: Reflection and Response
Regarding the ZTE Incident, Professor Wang Dongsheng, Director of the CPU Center of the National Laboratory for Information Science and Technology of Tsinghua University believes that the main problems exposed by the ZTE event are insufficient innovation of core technologies and source technologies. The Chinese semiconductor industry is like Chinese football, and it is also compared with international advanced enterprises. There is a big gap. The entire semiconductor industry chain in China does not have the right to speak. There should be a master plan at the national level, and increase investment in basic research and scientific research personnel. It should not be arrogant, but it cannot be arrogant.
Tao Yaodong, vice president of Corporate Security Group 360, believes that the ZTE incident should not rise to ethnic or shame. ZTE received several requests from the US Department of Commerce for rectification prior to ZTE, which was equivalent to making mistakes again. The incident should be viewed calmly.
Dr. Han Yinhe, a researcher at the Institute of Computing Technology of the Chinese Academy of Sciences, pointed out that the ZTE incident revealed that although China's complete machine and application industry is developing rapidly, there is a lack of infrastructure technology in the information industry, especially the chip industry. Support for basic research on semiconductor chips should be strengthened. The support mechanisms for enterprises need to be re-examined. Reliance on national projects has created many parasitized companies that rely entirely on national projects for blood transfusion. The lack of national investment has resulted in limited resources and disorderly competition. In addition, talent training is also critical. Research output in the field of chips With its own characteristics, we should pay attention to differences.
Dr. Wang Shijiang, deputy secretary-general of the China Semiconductor Industry Association, pointed out that at present, China-US trade disputes and ZTE events once again exposed the shortcomings of China's IC industry. Where future industrial policies go, they need to be given special attention. The most crucial point is to maintain policy. Stability and continuity, and second, to maintain the implementation of policies in place, most of the past industrial policies are concentrated in the design and manufacturing links, and should be supported in all aspects such as equipment and materials.
Stones from other hills: What experiences can you learn from Japan, Korea, and Taiwan in developing semiconductors?
Zhi Zhixin, a chip engineer who once worked in the IBM chip division, gave his opinion.
In his opinion, the development of the semiconductor industry in Taiwan first benefited from Government policies and talent introduction The most famous one is Tsung-Mou Zhang, the founder of TSMC. He was originally a senior director of Texas Instruments Semiconductor. He has already been an important figure in the United States. After returning to Taiwan, he established and expanded TSMC through his personal contacts and experience. Talent introduction is the core.
At the same time, the development of TSMC also depends on Development of peripheral industries , Including the supply of equipment, such as a series of upstream supplies such as Keguang, testing equipment, wafer suppliers, and a series of industrial chains such as chip companies and foundry factories.
Zhao Zhixin pointed out that another very important lesson is to Forming an industrial chain , Ziguang Group is now doing a very good job, first is the acquisition of Spreadtrum, to solve the demand for chip production, then the Yangtze River storage, began to enter the production.
'Market operation It is a shortcut, but this shortcut will be blocked by other countries. At present, domestic capital has not been very successful in acquiring a fab. Last year it successfully acquired a very small Italian factory, but the core of real technology is not easy to transfer. So we need to have our own R&D, only if our R&D is strong enough to have a real technology transfer.
In addition, he said that other countries have subsidies for universities in terms of chips and tapeout testing. Taiwan universities can test chips for free, the Canadian government gives about 2/3 discount on chips, and schools pay for research funding. 1/3. Europe also has a lot of university and fab research and development which is directly supported by the EU. Our chip development talents need to be cultivated by themselves. Therefore, it is very important to give more research support to universities. Talent must not rely on imports.
'In addition, domestic universities also need to strengthen cooperation, many universities have very expensive equipment, these devices can also share or slightly charge some maintenance costs, both to reduce total investment, but also can make full use of equipment resources,' said Zhao Zhixin.
Wang Shijiang mentioned that China began to pay attention to semiconductor layout from the 1950s, but missed the gold development period in the 1960s and 1970s. Most of the mainstream companies in the semiconductor industry rose in the 1960s and 1970s.
The main prototype of the semiconductor industry in Japan developed rapidly in the 1970s. Early chips were bought in the United States. However, Texas Instruments and Motorola all began to manufacture complete machines. As a result, Japanese companies did not compete with American companies in terms of complete machines. Researching chips. Most companies in Japan now do chips and systems. In the 1970s and 1980s, they accumulated technology. However, after China missed this wave, it will wait until 2000 to start again. The market has already been occupied, and the first place has taken over. The vast majority of the market, the second place to drink a little soup, the third place is immortal. It is very difficult to get back now after falling behind. But in emerging areas, such as artificial intelligence and blockchain chip research and development has an advantage, should catch Live opportunity to develop. '
Han Yinhe pointed out that at the time of Japan's revitalization, there were several concrete measures. First, it caught up with the semiconductor industry, especially the storage market. Second, Japan introduced the policy of cooperation between the government and the people, and the country set up a research institution specifically. After setting the DM as the main attack direction, after the technical research came out, it was all transferred to the investment alliance enterprise. The company introduced its own higher-order products and gradually iterated, which enabled Japan to complete the technical catch-up in the DM field.
Is chip development dependent on the government or on the market?
Cheng Zhenlin, an IC fund investor, is a firm market faction. He is not particularly optimistic about the policy. He believes that Japan’s policy is based on the country’s national conditions and cannot copy.
Cheng Zhenlin added that 'the market not only means full competition, but also includes market-oriented thinking. The current problem is not that policies are not in place, that marketization is not in place, and that policies for technological innovation should go further apart from basic research. , As long as it is related to the industrialization of products, they should all be solved with market-oriented thinking. The incentive mechanism, the assessment mechanism, the mechanism for circulating funds, and the project selection mechanism should all be market-based so that they can do a good job.
Wang Shijiang put forward different views. He pointed out that at least in the traditional chip market still needs some guidance. The government has some things to do to improve. Of course, the chip industry is a highly international and highly marketable industry, and should not be too dependent on it. In the government, only through competition can the products be introduced into new directions. At the same time, he suggests that in some emerging areas governments should not intervene too much and allow capital to play a major role.
Zhao Zhixin believes that markets and policies are very important, but industrial policies must be tilted because semiconductors are an extremely monopolistic industry. The barriers to entry are quite high. The establishment of a world-class chip factory requires close to $ 2 to $ 5 billion, market access and R&D. The cost is very high. If there is no national capital participation, the semiconductor industry is struggling. When capital is supported until the company can gain a firm foothold in the market, it is possible to realize real profits.
Han Yinhe pointed out that through researching the development of the information industry in China in the past 20 years, the personal computer market ranks first in the world, the mobile phone ranks first in the world, and the internet ranks second in the world. However, the development of these industries did not drive the rise of chips. The enthusiasm of the chip is not high, and its core chips and key components basically rely on trade configuration, which means that the market is not a panacea, but it shows that the chip is a category with low profit, long cycle and high risk. According to the profit-seeking mechanism of the market, it will not The priority was given to development. At the same time, the case of the rise of the semiconductor industry in Japan and South Korea at the same time also showed that there must be strong national support to achieve the rise of the post- national semiconductor industry.
Is the market for technology feasible?
Cheng Zhenlin believes that the market is a commercial activity, and technology is not a pure business. Market-for-technology is inherently unequal, and business is business. High-speed rail is a very special case that cannot be used as a typical argument in the high-tech field. If capacity is not improved, Every time I get a copy, I can't get high-tech.
Wang Shijiang put forward the opposite view. He believes that there are many benefits to the market for technology exchange. By introducing more companies into the country to build factories, as long as factories come in, they have to hire people, train people, and import raw material manufacturers accordingly. This can form an industry. With the agglomeration effect, supporting facilities will be continuously improved. After the formation of industrial agglomeration, localization of equipment and materials can reduce costs, and can create competitive advantages compared with foreign companies. Of course, technology cannot rely entirely on external markets.
summary:
According to the First Financial Report, China's chip market accounted for more than 50% of the global market, some chips accounted for 70% to 80%, and 90% relied on imports. Domestic chips can only supply about 8%. Only in 2016, China imported chips. The total amount is close to 1.5 trillion yuan, nearly double the amount of crude oil imported in the second place. China has shortcomings in chip design, manufacturing, etc., especially in the manufacturing sector is relatively weak, part of the core technology, the key equipment Not fully grasped.
The ZTE event reflects the dilemma of China's semiconductor industry. How to deal with the situation requires all parties to work together. The government should issue corresponding industrial policies and support plans, increase investment in basic research and the introduction and training of scientific research personnel, and increase the number of scientific research personnel. Treatment; Increase financial support, establish a special fund to promote the development of the industry, give full play to the role of the market, and create a good environment for enterprise development. For enterprises, we should strive to improve our core technology research and development capabilities, and break through the technological bottlenecks.
Institutions of higher learning and research institutes should perfect the talent assessment mechanism and vigorously cultivate talents in the chip research field. Purely relying on market mechanisms can not save the chip industry, but relying on the government alone is difficult to stimulate innovation dynamism. Market-for-technology is not always a long-term solution. Core technologies Breakthroughs in R&D need self-reliance.
