How big is the gap between the development level of China's chip industry and the United States? Eschips perspective
How big is the gap between the development level of China's chip industry and the United States? Eschips perspective
Due to the Sino-US trade war, people's attention to chips has risen sharply.
A chip, also known as a microcircuit, is a silicon chip that contains an integrated circuit inside. To make such a silicon chip, it takes nearly 5,000 processes to arrange hundreds of millions of transistors precisely on a fingernail-sized silicon chip. Chip design and manufacturing techniques are constantly iterated, and each iteration has a profound impact on the landscape of the semiconductor industry.
Since the 1950s, the integrated circuit industry in mainland China has experienced more than 60 years of development, but "lack of core and less soul" has been a major dilemma of industrial development. Since last year, the chip and operating system "cut off" incidents suffered by ZTE and Huawei have once again highlighted this dilemma.
Therefore, we cannot help but ask: What is the status quo and level of China's development? What is the biggest bottleneck we are facing? How far is there to go to break through the siege and become self-reliant?
We had a conversation with Mr. Ni Guangnan, an academician of the Chinese Academy of Engineering and a researcher at the Institute of Computing Technology of the Chinese Academy of Sciences. Text, video there is always a suitable for you.
Be on the alert
Xiake Island: Some people say that the number of propaganda reports on the promotion of innovation is not as profound and effective as the warnings of the "ZTE incident" and "Huawei incident". How do you view these two events?
Ni Guangnan: Before the "ZTE incident" and "Huawei incident" did not happen, many people think that chips are very ordinary electronic components, which are directly bought from the market. But in fact, chip technology is the commanding height of modern information technology, related to national competitiveness and information security.
The occurrence of the ZTE and Huawei incidents has made us realize the extremely significant value of chip technology and the chip industry, and formed a situation where the whole people are spontaneously concerned about the development of China's chip technology and the progress of the chip industry. In this sense, the above events, like a "wake-up needle," have a positive side.
Weak link
Xia Ke Island: After years of development, how is the development level of China's chip industry? How much of a gap is there compared to the United States? In what ways?
Ni Guangnan: In terms of chip design, China has done a good job. China has the world's largest number of chip design companies, chip design level is also ranked second in the world, after the United States. We have designed a number of excellent products, and the CPU (central processing unit) chip used in the Sunway Taihulight supercomputer, which has been ranked first in the world for several years, is a typical example. The CPU chip performance of domestic mobile phones, computers and servers has basically reached the international advanced level.
Even so, China still has a "short board" in the field of chip design tools. Chip design needs to rely on electronic design automation tool (EDA), which enables the designer to use the computer to compile, simplify, divide, synthesize, optimize, layout, routing and simulation, and finally complete the chip design. However, the software service is mainly provided by three US companies.
In the field of chip manufacturing, including manufacturing processes and manufacturing equipment, China's capabilities are still weak. Chip manufacturing sounds like traditional manufacturing, but in fact its manufacturing process and equipment are far more sophisticated and complex than the latter. Specifically, its processes include photolithography, etching, ion implantation, film growth, polishing, metallization, diffusion, oxidation... And so on.
Corresponding to the above manufacturing process, there are more than 200 kinds of key manufacturing equipment, including lithography machines, etching machines, cleaning machines, cutting and thinning equipment, sorting machines and other processes required for diffusion, oxidation, cleaning equipment. The manufacturing technology requirements of each kind of equipment are very high, the manufacturing difficulty is very high and the cost is very high.
At present, in the field of chip manufacturing, the leading enterprises are mostly from the United States, Japan and other countries, and 80% of the equipment of Chinese chip manufacturers needs to be imported from abroad.
With these advanced equipment is far from enough, but also to open production lines, develop business plans. Plant construction, equipment installation and commissioning often take 2-3 years, which means that chip manufacturers have to make judgments about market demand in advance. Chip manufacturing technology is constantly iterated and updated, and it is still unknown whether the previous equipment and production lines can meet the market demand when they are actually put into production. If the newly built production line is not fully realized, the previous investment will be at significant risk.
In addition, the materials required for chip manufacturing are also heavily dependent on imports, and some materials such as photoresist need to be imported. The overall scale of the domestic semiconductor material industry is small and the technical level is low, and the sales scale of domestic materials accounts for less than 5% of the global proportion, which is still a big gap compared with developed countries.
Xiakoshima: In many industrial fields, we have successfully achieved leapfrog and catch up, such as high-speed railways and household appliances, but in the field of integrated circuits, this scene has not been able to appear. Why?
Ni Guangnan: From the historical point of view, we started late. In 1947, the United States Bell Laboratories invented the semiconductor point-contact transistor. It was not until 1956 that China successfully made the first silicon single crystal. Given that early computers were made of discrete components (tubes, transistors, etc.), China could follow foreign computer technology back then.
With the development of integrated circuits to the era of large-scale integrated circuits, China is obviously weak to catch up due to the lack of support from the integrated circuit industry. However, the world's chip technology has advanced by leaps and bounds. So far, although we have internationally renowned chip design and chip manufacturing companies such as Huawei hisilicon and SMIC, in terms of overall industrial development, compared with the international advanced level, we may need one or two decades to catch up.
The late start is an important historical reason, and the deviation of the leading concept leads to insufficient innovation promotion is a practical reason. In the past, in many respects, we hoped to solve the problem in a more convenient way, the so-called "it is better to buy than to build, and it is better to buy than to rent". Practice has proved that core technology cannot be bought. The "short board" of China's chip industry ultimately needs the Chinese people to constantly promote innovation through hard work and catch up step by step.
Breakthrough
Xiake Island: In the face of internal "short board", external blockade and other problems, how can China's chip technology and industry achieve a breakthrough?
Ni Guangnan: From the perspective of supply chain security, once a certain link of the supply chain is "cut off", it will make the entire industry fall into passivity. In this case, it is necessary to focus on key core technologies to break through the weak links in the supply chain.
The development of key core technologies is often a process from scratch, from small to large, and there must be a "bench to sit ten years cold" ideological preparation.
The maturity of core technology is also inseparable from the support of the market, the newly developed core technology and products if there is no chance to trial and error in the market, there is no virtuous cycle of entering the market, whether it is software or hardware, it is difficult to reach the level of user satisfaction.
We stress that government procurement should support domestic independent research and development of chips and software, should highlight the requirements of network security, so that domestic independent research and development of chips and software have the opportunity to enter the market application, to be tested and tempered, only in this way, can continue to mature.
From the perspective of industrial clusters, we can refer to the industrial cluster model of Silicon Valley. In Silicon Valley, there are first-class research institutions and universities, many entrepreneurs, and abundant venture capital, which form a close cooperation network and realize the strong combination of talent, technology, capital and management.
Beijing's Zhongguancun has the potential to become a high ground for chip technology and industrial clusters. It is advisable to make full use of the surrounding scientific research institutes and human resources, coupled with the support of national policies and capital, coordinate funds and R & D resources, strengthen the cooperation between industry, university and research, and make a high-tech industrial cluster project with Chinese characteristics.
The investment cycle of the development of large-scale software or chip industry is often measured in one or two decades, and the capital investment is very large, and it is difficult to achieve a rapid breakthrough simply by the market and enterprises. Because, the general enterprise will have pressure, such as listed companies to produce statements, to make thick profits, to prove their performance to shareholders.
Therefore, it is necessary to carry forward the advantages of the combination of the national system and the market economy, such as the formulation of some long-term plans and development plans at the national level. At the same time, we should actively introduce supporting support from the financial market, such as the support of venture capital in capital, encourage enterprises to invest in research and development and production, and pay attention to balancing short-sighted behaviors that may occur in the process of research and development.
In general, the development of the integrated circuit industry, to have long-term ideological preparation and investment, can not expect to get returns in just a few years, the real development of the integrated circuit industry, I am afraid that it will take a decade or two, we must have the determination, but also have the determination, to make up the industry "short board", and firmly insist on doing it.
Due to the Sino-US trade war, people's attention to chips has risen sharply.
A chip, also known as a microcircuit, is a silicon chip that contains an integrated circuit inside. To make such a silicon chip, it takes nearly 5,000 processes to arrange hundreds of millions of transistors precisely on a fingernail-sized silicon chip. Chip design and manufacturing techniques are constantly iterated, and each iteration has a profound impact on the landscape of the semiconductor industry.
Since the 1950s, the integrated circuit industry in mainland China has experienced more than 60 years of development, but "lack of core and less soul" has been a major dilemma of industrial development. Since last year, the chip and operating system "cut off" incidents suffered by ZTE and Huawei have once again highlighted this dilemma.
Therefore, we cannot help but ask: What is the status quo and level of China's development? What is the biggest bottleneck we are facing? How far is there to go to break through the siege and become self-reliant?
We had a conversation with Mr. Ni Guangnan, an academician of the Chinese Academy of Engineering and a researcher at the Institute of Computing Technology of the Chinese Academy of Sciences. Text, video there is always a suitable for you.
Be on the alert
Xiake Island: Some people say that the number of propaganda reports on the promotion of innovation is not as profound and effective as the warnings of the "ZTE incident" and "Huawei incident". How do you view these two events?
Ni Guangnan: Before the "ZTE incident" and "Huawei incident" did not happen, many people think that chips are very ordinary electronic components, which are directly bought from the market. But in fact, chip technology is the commanding height of modern information technology, related to national competitiveness and information security.
The occurrence of the ZTE and Huawei incidents has made us realize the extremely significant value of chip technology and the chip industry, and formed a situation where the whole people are spontaneously concerned about the development of China's chip technology and the progress of the chip industry. In this sense, the above events, like a "wake-up needle," have a positive side.
Weak link
Xia Ke Island: After years of development, how is the development level of China's chip industry? How much of a gap is there compared to the United States? In what ways?
Ni Guangnan: In terms of chip design, China has done a good job. China has the world's largest number of chip design companies, chip design level is also ranked second in the world, after the United States. We have designed a number of excellent products, and the CPU (central processing unit) chip used in the Sunway Taihulight supercomputer, which has been ranked first in the world for several years, is a typical example. The CPU chip performance of domestic mobile phones, computers and servers has basically reached the international advanced level.
Even so, China still has a "short board" in the field of chip design tools. Chip design needs to rely on electronic design automation tool (EDA), which enables the designer to use the computer to compile, simplify, divide, synthesize, optimize, layout, routing and simulation, and finally complete the chip design. However, the software service is mainly provided by three US companies.
In the field of chip manufacturing, including manufacturing processes and manufacturing equipment, China's capabilities are still weak. Chip manufacturing sounds like traditional manufacturing, but in fact its manufacturing process and equipment are far more sophisticated and complex than the latter. Specifically, its processes include photolithography, etching, ion implantation, film growth, polishing, metallization, diffusion, oxidation... And so on.
Corresponding to the above manufacturing process, there are more than 200 kinds of key manufacturing equipment, including lithography machines, etching machines, cleaning machines, cutting and thinning equipment, sorting machines and other processes required for diffusion, oxidation, cleaning equipment. The manufacturing technology requirements of each kind of equipment are very high, the manufacturing difficulty is very high and the cost is very high.
At present, in the field of chip manufacturing, the leading enterprises are mostly from the United States, Japan and other countries, and 80% of the equipment of Chinese chip manufacturers needs to be imported from abroad.
With these advanced equipment is far from enough, but also to open production lines, develop business plans. Plant construction, equipment installation and commissioning often take 2-3 years, which means that chip manufacturers have to make judgments about market demand in advance. Chip manufacturing technology is constantly iterated and updated, and it is still unknown whether the previous equipment and production lines can meet the market demand when they are actually put into production. If the newly built production line is not fully realized, the previous investment will be at significant risk.
In addition, the materials required for chip manufacturing are also heavily dependent on imports, and some materials such as photoresist need to be imported. The overall scale of the domestic semiconductor material industry is small and the technical level is low, and the sales scale of domestic materials accounts for less than 5% of the global proportion, which is still a big gap compared with developed countries.
Xiakoshima: In many industrial fields, we have successfully achieved leapfrog and catch up, such as high-speed railways and household appliances, but in the field of integrated circuits, this scene has not been able to appear. Why?
Ni Guangnan: From the historical point of view, we started late. In 1947, the United States Bell Laboratories invented the semiconductor point-contact transistor. It was not until 1956 that China successfully made the first silicon single crystal. Given that early computers were made of discrete components (tubes, transistors, etc.), China could follow foreign computer technology back then.
With the development of integrated circuits to the era of large-scale integrated circuits, China is obviously weak to catch up due to the lack of support from the integrated circuit industry. However, the world's chip technology has advanced by leaps and bounds. So far, although we have internationally renowned chip design and chip manufacturing companies such as Huawei hisilicon and SMIC, in terms of overall industrial development, compared with the international advanced level, we may need one or two decades to catch up.
The late start is an important historical reason, and the deviation of the leading concept leads to insufficient innovation promotion is a practical reason. In the past, in many respects, we hoped to solve the problem in a more convenient way, the so-called "it is better to buy than to build, and it is better to buy than to rent". Practice has proved that core technology cannot be bought. The "short board" of China's chip industry ultimately needs the Chinese people to constantly promote innovation through hard work and catch up step by step.
Breakthrough
Xiake Island: In the face of internal "short board", external blockade and other problems, how can China's chip technology and industry achieve a breakthrough?
Ni Guangnan: From the perspective of supply chain security, once a certain link of the supply chain is "cut off", it will make the entire industry fall into passivity. In this case, it is necessary to focus on key core technologies to break through the weak links in the supply chain.
The development of key core technologies is often a process from scratch, from small to large, and there must be a "bench to sit ten years cold" ideological preparation.
The maturity of core technology is also inseparable from the support of the market, the newly developed core technology and products if there is no chance to trial and error in the market, there is no virtuous cycle of entering the market, whether it is software or hardware, it is difficult to reach the level of user satisfaction.
We stress that government procurement should support domestic independent research and development of chips and software, should highlight the requirements of network security, so that domestic independent research and development of chips and software have the opportunity to enter the market application, to be tested and tempered, only in this way, can continue to mature.
From the perspective of industrial clusters, we can refer to the industrial cluster model of Silicon Valley. In Silicon Valley, there are first-class research institutions and universities, many entrepreneurs, and abundant venture capital, which form a close cooperation network and realize the strong combination of talent, technology, capital and management.
Beijing's Zhongguancun has the potential to become a high ground for chip technology and industrial clusters. It is advisable to make full use of the surrounding scientific research institutes and human resources, coupled with the support of national policies and capital, coordinate funds and R & D resources, strengthen the cooperation between industry, university and research, and make a high-tech industrial cluster project with Chinese characteristics.
The investment cycle of the development of large-scale software or chip industry is often measured in one or two decades, and the capital investment is very large, and it is difficult to achieve a rapid breakthrough simply by the market and enterprises. Because, the general enterprise will have pressure, such as listed companies to produce statements, to make thick profits, to prove their performance to shareholders.
Therefore, it is necessary to carry forward the advantages of the combination of the national system and the market economy, such as the formulation of some long-term plans and development plans at the national level. At the same time, we should actively introduce supporting support from the financial market, such as the support of venture capital in capital, encourage enterprises to invest in research and development and production, and pay attention to balancing short-sighted behaviors that may occur in the process of research and development.
In general, the development of the integrated circuit industry, to have long-term ideological preparation and investment, can not expect to get returns in just a few years, the real development of the integrated circuit industry, I am afraid that it will take a decade or two, we must have the determination, but also have the determination, to make up the industry "short board", and firmly insist on doing it.
