After Huawei encountered TSMC's supply cut, Ren Zhengfei once said that although HiSilicon can design the world's top chips, the domestic chip manufacturing level cannot keep up, and it cannot be built for the time being, and Huawei cannot make products and manufacture chips.
This is the most real status quo of the domestic chip market, and there are huge shortcomings in the manufacturing industry in the field of heavy assets. As the most technologically advanced and largest foundry enterprise in the mainland market, according to the official public information of SMIC, the highest process it can mass-produce at this stage is the 14nm process.
On the surface, there is still a big gap between SMIC and TSMC, after all, the latter has been updated to the 3nm process. But the gap in process does not represent a gap in technology.
Under the leadership of Liang Mengsong, the rapid rise of SMIC in recent years is obvious to all, in just five years, it has iterated from 90nm to 14nm, and its update efficiency is far more than that of TSMC and Samsung.
The main reason why it is impossible to go further in cutting-edge processes is the lack of EUV lithography machines. Although SMIC had ordered EUV lithography machines from ASML in 2018, these devices have not yet arrived under the intervention of the United States.
In contrast, TSMC, which has the right of first refusal, has more than 80 EUV lithography machines in its hands, which is one of the main reasons why TSMC is firmly positioned as the world's largest foundry.
According to normal logic, if SMIC wants to break the current chip pattern, it should increase investment in advanced processes to overcome the difficulty of equipment shortage.
Unexpectedly, SMIC recently announced that it will invest $7.5 billion to build a 12-inch fab production line in Tianjin, focusing on 110nm to 28nm chip processes, with a planned monthly production capacity of 100,000 pieces.
It should be known that SMIC's revenue in the first half of 2022 was 24.592 billion yuan, but now it has invested 7.5 billion US dollars to expand the mature chip process, which can be said to be a "big deal".
In fact, since last year, SMIC has invested in expanding mature chip processes four times, with a cumulative investment of more than 170 billion yuan. In the eyes of the outside world, these moves by SMIC are equivalent to abandoning advanced processes and betting all the "treasures" on mature processes, and the gap between SMC and TSMC may also become larger and larger.
However, from the current situation, the result is just the opposite, and SMIC's bet on the "big hand" of the mature chip process seems to be the right bet.
According to the information released by the International Semiconductor Association, the global lack of cores has been alleviated, especially the advanced process technology, affected by the decline in the global smartphone market, cutting-edge chips are not lacking, but there is also a serious overcapacity.
For example, Qualcomm, MediaTek, etc., have cut orders for advanced process chips, while Intel and Apple have postponed the release of 3nm chips. Even TSMC plans to abandon the N3 process and even shut down some EUV lithography machines to reduce the cost of advanced processes.
In contrast, the mature process, its market demand is still strong. Although the performance of mature process chips is not as good as that of advanced processes, the scope of application is wider, such as smart home appliances, aerospace, 5G base stations, artificial intelligence, automobiles and many other key areas, all of which are chips above 28nm, and the market application rate is as high as 70%.
Obviously, SMIC's decision to expand the mature process is correct, and it is of great significance to the development of domestic chips.
As Wu Hanming, an academician of Chinese engineering, said: Compared with conquering EUV equipment and laying out the high-end chip market, improving the localization rate of mature chips and doing a good job in system optimization is the top priority, and only by first standing firm in the low-end market can we better impact the high-end.
According to public information, SMIC's yield in 28nm/14nm mature chips has reached the level comparable to TSMC, and its production efficiency, number of customers, order types, etc., have reached the world's first position.
It is worth emphasizing that SMIC has not only achieved a leading position in 28nmSoc chips, but also achieved a leading position in the BCD process field.
The BCD process is different from the Soc chip, it is a monolithic integrated process, which simply means that three different components are concentrated on one chip, mainly used in automotive, communications and other industries.
Up to now, SMIC's BCD process has exceeded 55nm, while TSMC and Samsung are still stuck at the 65nm node, and ST Ishizoectronics is only 90nm.
Today's mature chip market can be said to be the general trend. SMIC not only laid its own industry position, but also laid a solid foundation for the next rapid development.
Of course, this also means that TSMC does not want to see the situation. The reason is clear, because TSMC has built a 5nm fab in the United States, under the provisions of the US Chip Act, subsidized companies will be banned from expanding production capacity and creating new advanced processes in the Chinese mainland market for the next decade. In other words, TSMC has lost its qualification to compete with SMIC in the mainland chip market.
In addition, with the oversupply of the high-end chip market, TSMC, which is hindered in the N3 process, will not be able to widen the gap between it and SMIC again.
You know, the mainland has the world's largest chip consumer market, for foundry companies, as long as the technology is excellent, it will not worry about orders at all. The continuous revenue support will undoubtedly accelerate the pace of SMIC's expansion of its commercial territory and the pace of technological breakthroughs and innovations to a large extent.
As for TSMC, the future should be bright, after all, it has huge advantages in foundry technology and production capacity. However, Liu Deyin and other executives had to keep up with the footsteps of the United States and bow their knees. It is a pity that the United States has not regarded it as its own person, but has tried to transfer its core technology.
Nowadays, with the continuous rise of domestic chips, TSMC seems to have gone, but it can only be said that it is the result of self-inflicted wounds.