We all know that as traditional silicon cores approach their physical limits, the more advanced the process, the higher the research and development costs, OEM costs, etc. will be. Therefore, many netizens have questions about advanced process chips. Who would buy such an expensive chip? In addition, TSMC has repeatedly postponed the mass production of 3nm, which also made netizens question. Is TSMC also worried about insufficient demand?
It was precisely because of TSMC's hesitation that Samsung took the lead and took the lead in achieving mass production of 3nm. But recently, TSMC finally announced that 3nm will launch a mass production plan, and Wafer Factory 18 will become the focus of TSMC’s 3nm production. It is reported that the total investment of Wafer Fab 18 has reached 422.22 billion yuan, which is huge and is expected to provide more than 11,000 jobs.
Regarding the issues that netizens are worried about, TSMC has officially "shown down", Liu Deyin has also expressed its position, and the demand for 3nm process is very strong. The implication is that TSMC has no worries about its 3nm foundry orders. At present, the yield rate of TSMC's 3nm mass production is basically the same as that of 5nm mass production. However, due to the higher logic density of 3nm, which has increased by 60%, at the same speed, 3nm can consume less power and perform more powerfully.
As the world's most advanced semiconductor technology, the 3nm process is of great significance to cloud computing, mobile devices and high-speed networks. This is also the main reason why TSMC is confident in the demand for 3nm. According to estimates by industry insiders, in the past five years, 3nm mass production will release US$1.5 trillion in terminal product value, and the development potential is quite large.
But having said that, the author thinks that Liu Deyin’s description of “very strong demand” is still exaggerated. Judging from the current market situation, there are no other major customers besides and Apple that can afford TSMC's 3nm production capacity of . According to foreign media reports, the main customer of TSMC’s 3nm process is Apple. In 2023, Apple's first product using TSMC's 3nm process will be the M2 Pro chip, followed by M3 and A17 chips.
It is undeniable that if you can have Apple as a big customer, you basically don’t have to worry about orders. After all, the size of the iPhone is the largest in the industry. But suffering from "Apple dependence" also carries great risks. First, after the release of the iPhone14 series, Apple's reputation declined, shipments decreased, and many supply chain companies had their orders cut off. After losing orders from Apple, TSMC had no plan B.
Secondly, the iPhone15 series has been exposed and will continue to increase differentiation. You must know that the differentiation policy of iPhone14 and iPhone14 Pro has already caused strong dissatisfaction among consumers. If the iPhone15 series becomes more expensive, consumers will completely give up on Apple. TSMC's bet on Apple is undoubtedly a "big gamble". It will either win or fail.
In general, TSMC has obviously had the intention of "pro-" the United States, and building a factory in the United States is the choice made by TSMC. The higher the reliance on US companies, the more likely it is to be suppressed or "harvested". The worst-case scenario is that TSMC's technology and talents will be robbed, and it will eventually return to the mainland in disgrace. But how can TSMC, which has lost its core competitiveness, compete with Samsung, , and Intel ? Personally, I think this move is a bit risky. What do you think of it? Welcome to leave comments and messages!
