In the context of the digital economy, chips have become the most widely used major components of electronic devices. With the landing of technologies such as 5G and the Internet of Things, the use of chips has continued to rise. However, due to the many chain reactions caused by the epidemic, chip production capacity and supply have been greatly affected, and the "lack of core" in the downstream industry has become a common problem in the past one or two years.
In such a situation, the role of chips has received unprecedented attention, the value of domestic semiconductor companies is also rising, even if it is on the verge of bankruptcy of related companies, there are a large number of people rushing to "take over".
Recently, Tsinghua Unigroup, known as China's "chip aircraft carrier" and valued at 300 billion yuan, issued the latest restructuring plan and is expected to carry out a high proportion of liquidation. It is understood that after the huge debt of Tsinghua Unigroup of nearly 200 billion yuan, Midea, Hikvision, Ali and other companies have expressed their willingness to lend a helping hand, but eventually spent Zhilu Capital and Jianguang Assets with a state-owned background.
As a result, the high prosperity of the chip track can be seen. However, realistically speaking, there is a certain gap between the domestic independent research and development of chips and the frontier, but fortunately, there are a number of "Chinese cores" that are coming forward and following, and are using their own actions to change the industry pattern.
Small chips, big dreams
It is no exaggeration to say that without chips, human civilization would have to go back at least 50 years.
From common electronic products such as mobile phones and computers, to vehicles such as cars and airplanes, and even spacecraft and space stations, all modern equipment can find the existence of chips.
Therefore, although the size of the chip is small, it brings together the most cutting-edge technology of human beings. Whether it is EDA tools and architectures at the design level, or lithography machines and process processes at the manufacturing level, each step is the condensation of decades of technology and experience.
From this point of view, we can find that in the field of chips or semiconductor integrated circuits, head companies often need decades of technology accumulation, and it is difficult to make breakthroughs in a short period of time, which is one of the reasons why the current domestic chip field cannot break the foreign technology blockade.
In the past year, the development of 5G, AI, autonomous driving and consumer electronics industries has greatly driven the growth of the chip market. According to the performance of the secondary market, the semiconductor development index shows a volatile upward trend, and the stock price changes of SMIC, Weier shares and other head listed companies also reflect this, which means that China's semiconductor industry is in a stage of rapid development.
However, due to the limitations of domestic chip technology, it has not yet formed an advantage in the design field. Taking the process process as an example, now TSMC, Samsung and other first-line manufacturers have mastered the 5nm, 4nm process technology, and the domestically recognized head enterprise SMIC is still hovering at 14nm. Although the latter has conquered the 7nm process, it is still not a short distance from TSMC, Samsung and so on.
Optimistically, in terms of chip design, we are only lagging behind in cutting-edge technology, not without the opportunity to catch up. But in the field of chip manufacturing, the domestic level is one or two decades behind the international highest level.
According to TrendForce research data, in the global semiconductor foundry market in the third quarter of 2021, TSMC's market share reached 53.1%, far better than other companies, and SMIC, although it ranked sixth in the world overall, had a market share of only 4.3%.
The technology and manufacturing sectors are both lagging behind, which has greatly affected the domestic downstream industry in the face of the impact of the epidemic. Not only is the supply-side production capacity constrained, but the gap in chips has further widened under the stimulation of the epidemic.
Since the beginning of this year, a number of new energy vehicle brands, including Xiaopeng Automobile and Weilai Automobile, have faced the problem of delayed delivery, and the shortage of chips is one of the main reasons;
In addition, because of well-known reasons, Huawei Kirin chip manufacturing can not be carried out, mobile phone chips are jointly controlled by Qualcomm and MediaTek, after the recent launch of Qualcomm Snapdragon 898 flagship chips, Motorola and Xiaomi in order to compete for the first hit "head broken blood flow".
Chip shortages have hindered the development of many downstream industries, which has also spawned more players to enter the chip manufacturing. Among them, there are not only new players like Cambrian who are trying to seek breakthroughs with the help of AI technology, but even downstream companies such as OPPO have begun to cross-border "core-building".
But under the hustle and bustle, why has the high-tech "Chinese core" not yet appeared?
"Chinese can't make a chip"?
This problem is also a long-standing haze in the industry, and some people even think that "Chinese can not make chips". But in fact, the failure of the "Chinese core" to appear in time is produced by the superposition of complex historical factors and current practical factors.
In the past, in order to break the blockade of computer technology in Western countries on China, school-enterprise cooperation became the trend of the times at that time, of which Peking University Founder, Tsinghua Tsinghua Unigroup and Lenovo with the background of the Chinese Academy of Sciences were the three main ones.
With the large-scale intrusion of foreign computer brands into the Chinese market, most independent brands are unable to compete. Moreover, many brands have abandoned the research and development of core technologies such as chips in the shallow cognition of "making better than buying", and transformed into "assembly plants", such as Lenovo.
Although Tsinghua Unigroup chose to develop chips, it went astray. According to incomplete statistics, Tsinghua Unigroup has more than 60 mergers and acquisitions in the past ten years, involving almost every corner of chip manufacturing. Unfortunately, Tsinghua Unigroup did not become China's "Samsung", but was dragged down by huge debts.
However, during this period, there are still some people of insight who have proposed that the core technology cannot be controlled by people, and the most cutting-edge technology cannot be bought through purchase. Therefore, in 2001, the "Ark One" was introduced, but due to the lack of a supporting operating system and software, the chip failed to play its due role.
Subsequently, the "Hanxin incident" also covered the industry, and even the national system had to be cautious in research and development investment. These factors have caused China to miss the opportunity to catch up with chip technology.
At present, although there are many entrants, it is still very difficult to independently produce a high-tech chip. In addition to the continuation of historical reasons to the present, production equipment also restricts the emergence of "Chinese core".
In chip manufacturing, the most critical is the silicon wafer and lithography machine, of which the domestic silicon wafer materials have now caught up with the international leading echelon, but the lithography machine is really difficult for domestic chip manufacturing enterprises, because the most advanced 13.5nm wavelength EUV lithography machine is not allowed to sell to China.
The reason why humans can develop smaller and smaller chips is mainly due to the continuous improvement of ultraviolet wavelengths. If you can't buy advanced lithography machines, you obviously can't create advanced chips.
That is to say, if you want to solve the problem of high-end chip production, you must first solve the production problem of chip production tools, and self-research has become the only hope. According to the existing news, the top domestic scientific research institutions such as the Chinese Academy of Sciences have entered the bureau, which is also one of the important layouts of domestic semiconductors.
In addition, the investment in research and development costs is small, which is also one of the reasons why China's chip manufacturing enterprises lag behind abroad.
In recent years, China's semiconductor companies have caught up with the momentum very quickly, and they have not spared investment in research and development, such as the upstream EDA enterprise Synopsys Technology, which accounts for more than 30% of the total revenue every year. According to the report of IC Insights, an authoritative semiconductor market research company, semiconductor R&D expenditure accounted for only 14.2% of the total global semiconductor revenue in 2020.
However, although the proportion is far above the average, due to the lack of domestic enterprises, the total R& D expenditure invested is not enough. In 2020, TSMC's R&D expenses will be about 25.1 billion yuan, MediaTek's R&D expenses will be about 11 billion yuan, while SMIC's annual R&D expenses will be only 1.08 billion yuan.
The chip industry is a capital-intensive industry, and only by increasing investment in the field of scientific research will the development of the technology field be faster. However, in the past two years, the National Integrated Circuit Industry Investment Fund has cumulatively driven 800 billion yuan of funds to invest in the semiconductor industry chain, and the high-tech "Chinese core" may be on the way. However, from an objective point of view, how to go about the domestic chip semiconductor industry?
Where is the opportunity for China's independent chips to catch up?
To be fair, if you want to fully realize the complete autonomy of domestic chips, you still need a certain amount of time and capital investment. Although there are already high-end chips such as the Loongson series and the Huawei HiSilicon series used in spacecraft, they still need a long time from commercialization and cannot meet the huge domestic consumer demand.
Public information shows that China is the world's largest chip consumer market, in order to cope with the increasingly large and diverse demand for use, in 2020 alone, China imported 350 billion US dollars worth of chips from overseas, accounting for more than 70% of the total domestic consumption of chips that year.
From this phenomenon, we can see that the chip field of non-cutting-edge process technology may become an opportunity for the domestic semiconductor industry to catch up.
Since the 7nm process, the impact of advanced process improvement on the performance of civilian intelligent equipment has become more and more limited. And the related industries involved in the Internet of Everything, the mainstream chips required are still focused on the mature processes of 28nm and 14nm, at least in the field of new energy vehicles, chips using low-thread processes cannot fully contribute to product performance.
Under such circumstances, domestic chip companies should seize such opportunities to gradually extend from low-end to high-end through volume advantages, providing the possibility of overtaking in the corner of China's semiconductor industry.
Secondly, the use of more advanced technologies such as AI, chip companies can meet more needs for downstream manufacturers, so as to achieve a breakthrough, such as the Cambrian known as the "first share of AI chips", which shows another possibility for the industry.
In addition, the field of chip packaging is becoming a new opportunity for the industry.
As the "last mile" of chip production, chip packaging has a decisive factor for the availability and stability of the product, and there are still many places to make a fuss about this aspect. At the 2021 World Semiconductor Conference, Mao Jun, an academician of the Chinese Academy of Sciences, proposed that heterogeneous integrated circuits can be used to help the domestic chip industry to the forefront.
The so-called heterogeneous integrated circuit is the same 3D system-level package, the silicon and non-silicon devices manufactured in different processes are integrated into a higher-level system, so that the secondary technology can also play a higher performance.
According to "Pinecone Finance", through heterogeneous integration, the chip performance of domestic companies with 40nm processes can already be comparable to 16nm chips. Through a more mature process to make a more advanced system, perhaps to a certain extent, can represent the future technology extension and development direction.
This view is not an isolated case, and Intel has previously said that heterogeneous integrated circuits are the future trend of packaging technology.
However, objectively speaking, the current Chinese chip packaging is still dominated by traditional technology, and there is also a certain gap with the international leading level. Although enterprises such as Changdian, Tongfu, Huatian and Jingfang have begun to accumulate technology through research and development and mergers and acquisitions, domestic advanced packaging revenue accounts for only 25% of total revenue, far below the global average of 41%, and it is necessary to continue to increase investment.
epilogue:
Semiconductors determine the future of manufacturing, but the current market challenges and opportunities coexist, from SMIC, Huawei HiSilicon, Silan Micro, etc., we can see the efforts of Chinese in the industry independence, but also see the resilience of China's chip industry in high dynamic market changes, looking forward to the early emergence of the high-tech "Chinese core".