Wang Ruchen/Wen
In the spring of 2022, the two major regions in Europe and the United States announced the new policies for the semiconductor industry that have been hyped endlessly since last year. The numbers are pretty scary.
First, the Chip Act of the U.S. Competition Act of 2022 shows that a chip fund will be created to allocate $52 billion to encourage private companies to invest in semiconductor production. In addition, 45 billion yuan will be authorized to improve the supply chain. Strengthen manufacturing.
Second, the European Commission's Chip Act is 43 billion euros (nearly $49 billion). It is said to support chip production, pilot projects and a new generation of chip factories to increase the EU's global chip production capacity share, resolve the global supply chain crisis, and enhance the competitiveness of the digital economy.
It seems that the mountain rain is coming.
However, combined with the current situation of the chip industry, the epidemic situation and the global digital wave, I really feel that if the two major regions only focus on the so-called "production capacity", the final huge investment will inevitably become a gimmick, and the actual results may be chicken feathers.
As for the chinese people who have imagined that China will suffer a blow, they are completely unaware of the Chinese opportunities hidden behind this trend.
Many people are concerned about the crisis in the global semiconductor industry supply chain since the epidemic.
Indeed, in addition to the consumer electronics and mobile phone giants being tortured, the core industry in Europe and the United States, the big manufacturers in the automobile industry, have been tested one after another. Some new car production lines have been stopped for a long time.
In order to compete for limited production capacity, the global chip market is rampant with holiday goods. Even the robbery came out. This was unimaginable in the past.
This must have stung Europe and the United States.
But we don't think the two chip bills are born out of the immediate supply chain crisis.
If only to solve this challenge, even if you invest $100 billion each to build a huge semiconductor factory, it will not help. After all, from the time of construction to mass production, at least two years, far from quenching the thirst. This does not include uncertainties such as upstream material/equipment supply, design links, process realization, and final yield.
Europe and the United States each emphasized how brilliant the semiconductor manufacturing capacity was many years ago, how high the global proportion was, and how low it is today. The logic is similar and synchronized.
It seems that "production capacity" has become the crux of today's supply chain crisis.
On the surface, this is indeed the case.
In the past two years, around the semiconductor production capacity, Europe and the United States have indeed attracted TSMC, Intel, Samsung, UMC, and Grolfonid. Even Continental's SMIC is fully capable. Those who don't know Foundry even wonder about the more than 100% capacity utilization rate of some factories. Of course, not only semiconductors, but also many other components.
There is indeed a gap in "production capacity".
But there's a structural topic here.
The pandemic has catalyzed the need for a variety of online and digital services. But first of all, we must first meet the services of production, living, office, learning, social governance and infrastructure. Logistics, communications, power consumption, the Internet of Things, etc. are more urgent. Cars that are 2C bulky parts have been restricted due to the epidemic. In 2020, 80% of the country's car sales fell significantly. Among them, the United States has been beaten back to 8 years ago.
This, combined with supply chain challenges with chips and other automotive electronics, has exacerbated the sense of crisis.
The automobile industry is precisely the pillar industry of developed regions such as Europe and the United States, and the value chain is very long.
The so-called "insufficient production capacity" has a deeper sense of the authorities and industries in Europe and the United States.
Therefore, the competition for TSMC's production capacity has almost become the most significant geopolitical topic since the epidemic. This is true in the United States, and it is also true in Europe.
Initially, there was no vaccine in Taiwan, and some even asked the authorities to blackmail the chip in exchange. How distorted the industry has been.
However, the distribution of TSMC's foundry products, such as it, is unlikely to completely turn its services to the dominant industries in Europe and the United States.
During the same period, a number of semiconductor manufacturing projects such as Chinese mainland were launched. Despite many twists and turns, even under the clampdown of the United States and Europe and the obstruction of the Taiwan region, a small number of strands still conveyed optimism to the outside world.
This has exacerbated anxiety in Europe and the United States. And it continues to spread.
But the crux of the matter is not "capacity" itself.
The real reason is much more complex. There are three levels that can clearly perceive:
1. The epidemic prevention and control in core countries in Europe and the United States has not been optimistic, which has delayed economic recovery and seriously affected the openness and efficiency of the core industrial supply chain. The supply chain crisis has evolved from the initial epidemic to a crisis generated by global geopolitics and the new Cold War.
2, the United States clamped down on China, delayed the development of the latter, but because China in the global 600 industry categories of more than 200 types of cattle ears, and many intermediate products are also the world's first, the United States not only suffered from it, but also paid a higher price. These include end devices embedded in various semiconductors.
In fact, the supply chain crisis has changed from the initial epidemic to a global geopolitical and industrial crisis under the new Cold War, which has become a man-made disaster.
3. Focusing on semiconductors, behind the capacity crisis, in fact, the semiconductor industry supply chain in the era of globalization has encountered the double delay of the epidemic and man-made disasters.
The other, on the other hand, is easy to overlook.
That is, when the mobile Internet trend is trapped by the demographic dividend and the traffic peaks, supply-side innovation, especially infrastructure and digital and intelligent upgrades covering many industries, has become the obvious science of this era.
It is also one of the world's greatest certainties for years to come.
The corresponding is a new era of computing that is larger, more diverse, more complex and imaginative. This is an era of ubiquitous computing and intelligence. Not only will chip demand increase geometrically, but chip diversity will also bring many opportunities.
This is a window of transition.
"Insufficient capacity", in addition to the lack of absolute volume, is also difficult to respond to the needs of diversity.
Europe and the United States to strengthen the "Chip Act", strengthen the layout of manufacturing and the entire chain, if you look at this dimension, it is reasonable. However, the actual excuse has become a competition for future global semiconductor production capacity and manufacturing power.
Of course, "insufficient capacity" and "supply chain crisis" are still beautiful rhetoric.
In this way, they can stir up public opinion, activate populism and industrial nationalism, and try to reconstruct the map of the global semiconductor industry and the future digital economy.
After all, the supply chain crisis since the epidemic, as well as the increasing cross-strait semiconductor manufacturing, or Asian semiconductor manufacturing (including South Korea, the Middle East, Japan, Malaysia, Israel, etc.), is indeed more active than other regions. The above excuse is indeed the most convenient.
Therefore, the two major bills were finally passed, and there is still its logic.
But in any case, this still masks the division of labor trends experienced by the American and European semiconductor industries many years ago: initially they were all IDM models dominated the world. There are many giants. This was followed by a spectacular horizontal division of labor. TSMC is a product of the latter. ASLLs, of course, are too. The same logic is shared in other segments of global electronics.
The horizontal division of labor is also a manifestation of industrial democratization. In addition to the distribution of industrial chains, there are also industrial clusters in regional countries and regions. Although the computer industry, the Internet, the underlying architecture of semiconductors, and key materials/equipment/tools all come from the United States and Europe (Japan, South Korea, and Taiwan also have a considerable part), Chinese mainland has irreplaceable advantages in semiconductor foundry and design in terms of market demand, systems, terminals, design, industry applications, scenarios, etc., and the degree of coupling is high.
In 2013, Zhang Zhongmou said that in the future, the mainland will influence the power of the semiconductor industry, including Huawei and Alibaba.
If you couple TSMC, Huawei, and Alibaba, it is obviously a powerful coordination system. If we further integrate the global upstream forces and the global design, the two sides of the strait can leverage a larger situation. At the same time, do not ignore the trend of the mainland's comprehensive layout of the semiconductor industry in recent years, from materials, equipment, tools, design to manufacturing, packaging and testing has been relatively complete. Although the whole is still inferior to Europe and the United States, with the synergy of the entire system, especially the data and application scenarios in the era of digital economy, there are also irreplaceable conditions in the world. The rise and surpassing of the mainland semiconductor industry is only a matter of time.
Honestly, our semiconductor industry has had too much sadness over the years. However, in the past two years, under the restraint of the United States, it has more self-confidence. The industry's tragic view is evolving into a new normal that is still firmly developed under the pressure. In other words, China has turned a blind eye to the clampdown and is not strange. This mentality is the most natural. The corresponding industrial synergy will also enter a conscious cycle.
The European and American semiconductor industries conform to the industrial division of labor and globalization, and eat up the most profitable industrial links. Many giants have influenced the entire value chain and were once "first-class citizens" of many regional markets. Especially the giants of the semiconductor industry. They once had a sacred face.
Nowadays, when other regions have adapted to the industrial division of labor and tried to gradually jump to the core of the value chain with the help of the Internet, mobile Internet, and digital economy waves, Europe and the United States do not recognize their own dominant division of labor system.
Because, in this process, the underlying structure of the past has been difficult to influence the entire value chain, let alone eat the cake.
The two major "chip laws" in Europe and the United States seem to be based on the excuse of "insufficient production capacity" and "supply chain crisis", but in fact they have the intention of reversing and reshaping the rules of the game in the global semiconductor industry.
You can see that behind the two major bills, there are semiconductor giants. On the U.S. side, Intel is pushing hard. Chairman Pat Kissinger even once called the board, and the company even used the Xinjiang topic to manipulate and join the authorities in order to maximize benefits. After all, in terms of manufacturing, it is still the first in the United States, a global giant. In the EU, of course, Infineon is very active.
Because, 52 billion US dollars, 43 billion euros, will be converted into subsidies.
In a so-called liberal capitalist society, the best way to promote the birth of such bills and get follow-up subsidies is to constantly clamor and play up the industrial crisis, especially in Taiwan, the mainland and even Asia. Populism is always a good business.
Two bills, two huge numbers.
People are worried that this will have a huge impact on TSMC. It is also feared that Chinese mainland will continue to be pushed to the end of the semiconductor industry.
It's a matter of worrying. Not to mention the general trend of the entire digital economy, just two bills say that even if it can land, it will take a long cycle, and there are many things to do behind it.
And my own judgment is that if you look at short-term results, you must be a chicken feather. Of course, there will be effects in the long term.
Because neither the American version nor the European version fully emphasizes the so-called "self-sufficiency" of production capacity.
In the short term, the two bills are more like an investment promotion behavior. It is in a specific period, with the help of policies and administrative instructions, forcing semiconductor manufacturing to land in europe and the United States.
While the United States appeased Intel, it constantly coerced TSMC and Samsung to invest in projects as soon as possible. Although the latter two are well-spoken, the actual promotion efficiency is not so high. While Intel cooperates with TSMC, it constantly quarrels with it.
Similarly, infineon, the core players under the European Union, are also cooperating with TSMC on the one hand, and on the other hand, they are worried that the bill and many subsidies will fall into the hands of outsiders.
Last year, for example, Infineon CEO Reinhard Glaus said that if the EU finances incentivize global chip-making giants to get involved, their biggest customers will still be foreign tech giants, not European companies.
This expression is clearly jealous. After all, TSMC, Intel, and Samsung's IC manufacturing footprint is much larger than it.
But this is certainly unstoppable. The landing of new factories by TSMC and Samsung in the United States is relatively clear. Continuing to land in the EU is probably not a thing. Last year, for example, TSMC Chairman Liu Deyin said that it was unrealistic for the European bill to increase production capacity. But then he changed his mind. Later, it was reported that Espresso electric could set up a new factory in Germany. Can eat subsidies, of course he changed his mouth. Intel has no message yet.
Therefore, it is difficult to stand up for the idea that the two major laws in Europe and the United States will impact TSMC or other foundry giants.
This is why I say that the bill landed, first of all, it is a business promotion. Quite "powerful" is the giant investment in the establishment of factories. This is the fastest and easiest to see.
Of course, it also has the most ecological effect. After all, once TSMC, Intel, and Samsung land, there will be a group of upstream supporting enterprises landing, thus forming an industrial cluster and creating growth space for players such as material equipment and design companies in the two regions. Of course, it will also benefit the digitization process in many industries. In the eu, automobiles, communications, medical, industrial automation, steel, aviation and so on will have support systems.
Certainly not a player who denies Europe and the United States themselves. They are certainly the main force of the bill's incentives.
You see the previous paragraph Pat. Kissinger's. They all met with Petsy. The EU gang is too.
However, in terms of chip manufacturing diversity, influence, global supply chain controllability and urgency, TSMC is indeed indispensable. It must be said that the manufacturing end is the form of an aircraft carrier.
Europe and the United States certainly have the intention of cultivating independent new giants. But it is impossible without a long process. Don't expect the United States under the federal system, the Europe under the covenant, the government to lead the chip manufacturing project, there will be a big success. Then democracy, then liberal capitalism, is enough for the decision-making process alone.
You should see that the EU emphasizes that it does not deliberately pursue complete self-sufficiency, which even violates market rules.
Of course, if there is an ideal synergy, Europe and the United States have a new type of IDM model space. Their elements are relatively complete.
This made me think of the third-generation IDM model that SMIC founder Zhang Rujing has been calling for in recent years. Also known as CIDM. Different from the vertical integration of early monolithic enterprises, "C" is Commune, which is shared and shared, and there is room for upstream and downstream value chain collaboration between participants. Of course, it can also include the government.
This is a relatively ideal model. Especially in the era of digital economy for anti-monopoly, interconnection and deepening of synergy, CIDM does have room for birth. Europe and the United States are also.
Of course, the actual landing is definitely very troublesome. Because, with the increase in the diversity of chip demand, many chip companies and technology giants that did not have direct competition in the past suddenly have similar faces. The concepts in which they are involved are increasingly similar. This means that even under administrative instructions, it is unimaginable to form a synergy, but to work together on the plate of an independent legal person. This is indeed not the era of division of labor in the past.
Unless the executive directive is tough and sustainable. But in reality it is impossible. After $52 billion and €43 billion, who can trust that European and American principals can play a central role in the future. Besides, these two numbers seem to be very large, and if you look at the entire industrial chain and ecosystem, it is actually only a small amount.
Be aware that incentives and subsidies are not the same thing as the subsequent huge operating expenses. Administrative directives can form policy guidance, but they cannot directly influence the success or failure of a market entity.
The EU's chip bill does provide an idea of industrial synergy, that is, the official leadership of the establishment of some pilot production lines to accommodate faster and more efficient innovation. Of course, production capacity is also given priority to supplying itself. This is very similar to the way the Internet industry plays. It is also similar to new manufacturing such as Xiaomi.
In fact, the semiconductor industry has no shortage of such types. Even some of the top colleges with stronger microelectronics majors have their own test lines.
I actually agree with the landing of this dimension under the migration of supply chains. It can not only continue the industrial division of labor system in the context of globalization in the past, but also inject local vitality into it.
Overall, the path should tend to be flexible: official-led small-scale projects, CIDM-like, foreign monk chanting, and so on.
The ultimate goal is to achieve production capacity, supply chain controllability, and even chip sovereign cloud, with the intention of geopolitical competition. Don't say that there is competition between Europe and the United States and Asia, even between Europe and the United States. They are in cahoots, and they are even more intrigued. Not only in the field of chips, but also in the field of the Internet and the entire digital economy. For example, don't look at the US government under the coercion of a group of its own giants to torture Asia, especially Chinese mainland technology companies, FB, Google, Microsoft have also been repeatedly tortured by the European Union.
This does have security considerations. In the era of digital economy, data sovereignty has landed, but the real formation of arch guards and barriers, the chip industry is indeed a core defense line.
Even so, those who render two major bills and two huge figures really do not have to worry about the Chinese mainland in the new round of competition will become the end of the industrial chain. Because the collision of the needs, capabilities, volume, diversity of big countries and chips and even the entire digital and intelligent trend contains no less space for innovation. No matter how much the semiconductor industry is globally respected, it has always been a tool attribute. As long as it is a tool, for Chinese, the new cycle is not a real thing.
In my opinion, once the two major laws in Europe and the United States are really landed, the production capacity and capacity they generate cannot rely on their own internal digestion. They must still rely on larger markets and scenarios. China's dominant elements will not disappear.
And the elements of commerce, industry, study, research and politics that we separated in the past have become a distributed intelligent pattern today.
Although the road ahead is still dark, with internal pressures intertwined with external clamps, I have always firmly believed that the core capabilities of the major semiconductor industry will be generated in the next 5 to 10 years. In fact, this ability is more based on the synergy of other elements. The so-called kung fu is outside the poem. This is also one of the charms of the digital economy era.