Electrical Equipment / Components & Telecoms

There is a news recently that TSMC intends to invest US$2.8 billion to expand the monthly output of 40,000 28-nanometers in the Nanjing plant. Do you think this is a good thing or a bad thing?

For example, a big V believes: "TSMC’s expansion in Nanjing will accurately target mainland chip manufacturing companies in 2023. Using the mainland’s water, mainland electricity, mainland policies, and cheap manpower from the mainland, we will accurately target the emergence of mainland chip manufacturing companies. Just die, so that the advanced manufacturing process helps the United States fully control the backward manufacturing process and use dumping to kill mainland competitors. In TSMC’s mind, mainland officials and people on the mainland are all fools."

Many people heard it, yes! The strategy of imperialism is like this. When you can't make it, you will not sell it to you. If you are about to make it, you will reduce the price and dump it. It is too vicious to let your investment go to waste. Resolutely oppose the conspiracy of the United States and Taiwan!

However, I need to tell everyone that this is only the first level of consideration, and it is easy for laymen to think like this without knowing it. If you learn more and reach the second level, you will find that the answer is just the opposite. It is a good thing for TSMC to expand production in the mainland. I have discussed with a number of professionals in the chip industry, including chip manufacturing, chip design, chip consulting, chip basic research, and formulating industrial policies. They all agree with me. In fact, in the eyes of the industry, this is simply a matter of common sense, and one can see at a glance.

Why is there such a big disagreement? The people in front would say that I and those who agree with me are too naive and have no experience in the industry. In other words, I and those who agree with me don't understand political struggles, and don't know that the United States will lay down black hands. In other words, I and those who agree with me support Taiwan independence.

Do you think I support Taiwan independence? This accusation is too ridiculous and not worth refuting. In fact, making such outrageous accusations shows that they did not understand rational analysis at all, so they used all kinds of absurd motives to explain. First of all, it must be stated that my friends and I have a common position: to support independent research and development under any circumstances, and to support China's technological self-reliance and self-reliance. This is not what I said recently, but when I first started surfing the Internet more than two decades ago. High-tech must be in our own hands instead of expecting others to show kindness. This is my consistent view. Therefore, if you intend to tell me the truth in this respect, it is totally unnecessary. I know and speak these truths by no means less than anyone else.

So, the really interesting question is, why do those of us who fully support technological autonomy think that TSMC’s expansion of production is a good thing? Isn’t it afraid that TSMC dumps chip companies in mainland China?

The short answer is: The bottleneck of China's chip industry lies in technology, not in the market. If you understand the state of China's chip industry, then this sentence is enough. If you lack understanding, then I will start talking.

A basic question is, does China master the technology of 28nm chip manufacturing? The layman will think that they have mastered it, and then think that Chinese companies are investing in 28 nanometers, and the expansion of TSMC will crush them. But in fact, the answer is that China has not mastered it.

You may be wondering, isn't SMIC already producing energy at 14 nanometers? Isn't 28 nanometers already settled? Yes, but these have a premise: the use of American technology. As long as the US technology is used, the US government may issue a notice at any time, prohibiting you from supplying supplies to a certain company, such as Huawei, otherwise it will get stuck on your neck.

SMIC is like this now and cannot produce chips for Huawei. If you think about it carefully, a Chinese company has to obey the management of the United States and not supply to another Chinese company. Why? It is because the technology is still in the American system. There is only one way to get rid of this situation: develop an Americanized technological system. My friends and I have had many articles discussing how to develop a de-Americanization technology system, such as recruiting talents from all over the world, uniting the world outside the United States, using open source as a strategy, and using economic leverage to encourage companies from all over the world to beautify and many more.

So you can see that China has no 28-nanometer technology to beautify. This is a very important node. If you remember this, your knowledge level will exceed 90% of the people.

So, if the United States completely cuts off its supply to China, what kind of chips can China make? A more technical answer is 90 nanometers. The reason is that the most advanced lithography machine that China has built solely on its own technology corresponds to 90 nanometers.

You may feel cold, because we are so weak. You may also think that, fortunately, there is still a bottom line of 90 nanometers, which will not be annihilated. But the more expert answer is that 90 nanometers are not sure, because this lithography machine has not been used in the manufacturing process. And even if the photolithography function can be solved, we still have many materials that are not well made, such as photoresist and hydrofluoric acid. Therefore, whether 90-nanometer chip production can really be done remains to be tested.

By the way, although the 90-nanometer lithography machine is far from the cutting-edge technology of ASML, it is after all the achievements of China's hard work for decades. Most countries in the world have no intention of developing an independent chip industry, and have given up this competition long ago, and China has been insisting on it. This is commendable.

Some people often think that China's chip industry was once very advanced, and they blamed it for abandoning research and development after the reform and opening up. In fact, this idea is completely wrong. China's chip industry has never been advanced, and has never interrupted research and development. It's just because I have invested too little and others have made too fast, so I haven't caught up. But this is by no means a reason to fantasize about an "advanced past." The correct attitude is to increase investment and strive to catch up.

Academician Wang Shouwu, one of the founders of China's semiconductor technology

After understanding these backgrounds, you will understand that China’s top priority now is to master independent 28-nanometer technology. This is the core research goal of the entire Chinese chip industry, and strive to achieve it within a few years. Once we have technology, we need to build production lines, increase the yield rate, and ramp up production capacity. These are all full of uncertainties and may take several years. In short, this is a daunting task.

Therefore, if an independent 28-nanometer production line is built, it will be the top priority of the entire country and is related to national security. Do you think such a national weapon will be overwhelmed by dumping? of course not.

First of all, the government will have countless ways to support it, from policies to funds to all kinds of methods you can't think of. Then, as long as any company has independent needs, it will think about protecting its supply chain security and will not put eggs in a basket, so it will definitely give it an order. In addition, China's wafer production capacity only accounts for 14% of the world's, while demand accounts for about half of the world's, and the gap between supply and demand is huge. In 2020, China will import nearly 380 billion U.S. dollars of chips, of which more than 90% are 28 nanometers or above. Therefore, the number of expansion of TSMC is not to say that it is 40,000 pieces per month, even 400,000 pieces can be accommodated.

In December 2019, the wafer production capacity in mainland China accounted for only 13.9% of the world's total. Source: IC Insight

Therefore, it is completely unfounded to think that TSMC’s dumping will crush China’s chip industry. As long as the government gives equal treatment to foreign capital and domestic capital, there is nothing to worry about. Such worries may be used in the automotive industry, home appliance industry, cosmetics industry or the like in conventional industries, but it does not apply to the chip industry. If you can't see the difference between them, it means you know too little.

There is another layer that many people don't know is that in the chip industry, the situation facing the chip design industry and the chip manufacturing industry is very different.

In terms of chip design, many good companies have emerged in China, such as Godson, Cambrian, and Huawei HiSilicon. Their bottleneck is indeed largely in the market. In the 2020 viewing video year-end show, Mr. Lin Baojun, the chief designer of the Beidou-3 satellite system, told the story of how Godson was used in Beidou. He strongly urged everyone to give more opportunities to use Chinese chips, so that they will improve and make progress when used. I totally agree with these, and I often recommend Chinese chips to everyone, criticizing those fools who regard Godson as Hanxin and then scold Godson for counterfeiting.

Beidou 3 adopts Loongson plus FLASH architecture to replace imported CPU and PROM

In terms of chip manufacturing, it's a different story. If the difficulty of chip design is very high, then the difficulty of chip manufacturing is extremely high. There are few companies in the world that can make chips, and there are even fewer Chinese companies. Only SMIC can barely make advanced manufacturing processes, but there is still a big gap between TSMC. Their bottleneck is technology, not the market. As long as there is a technological breakthrough, there will be markets.

Process roadmap of major global foundries

Knowing these backgrounds, you will know how to understand TSMC's expansion. It has no impact on China's independent 28-nanometer production line, because the first has not yet been made, and after the second has been made, there will be no external forces that can impact this country's important weapon. It will have an impact on other companies' existing production lines under the US system, but this is originally a normal market competition. In fact, SMIC, Huahong, etc. are also expanding production, and more.

Now you can understand that there is almost no harm in TSMC's expansion. With this foundation, you can calmly see what benefits it has.

The most basic benefit is the production of chips. The layman may not realize that there is a shortage of chips all over the world. The first reason is the increase in demand for chips driven by new energy vehicles, and the second is that the political suppression of China by the United States has led to an increase in the insecurity of global companies in the supply of chips. In the past, there was always zero inventory, but now they are hoarding goods for three months. Suddenly the demand has increased a lot, and there is a shortage.

Recently, a shortage of automotive chips in the United States has caused large-scale car production to stop. In 47 days, Biden held two meetings at the White House, holding up a chip at a time and a wafer at a time, calling for a solution to the shortage of chips.

At the "Semiconductor Summit" held on April 12, 2021, Biden held up a wafer and claimed that "the United States is lagging behind in semiconductor manufacturing and R&D" and "we need to strengthen the domestic semiconductor industry and supply chain."

So, why do you oppose TSMC's expansion in China? Do you hope it will expand production in the United States to help the United States solve the chip shortage?

From a Chinese perspective, do you support the development of Chinese chip design companies? Of course support. So, should they produce their own chips? Of course it needs. Who will produce? Of course it is not themselves, but chip manufacturing companies such as TSMC, SMIC, and Hua Hong. So, if you oppose TSMC’s expansion, wouldn’t it be the neck of its own chip design company?

Someone may want to say, I support them to find SMIC or Hua Hong instead of TSMC. Then the problem is that some advanced manufacturing processes cannot be produced by SMIC and Hua Hong, only TSMC, what do you do? Someone might say again, it would rather not produce, than let TSMC make money. So, what should those chip design companies do? Should they go bankrupt?

Someone might say, yes, bankruptcy should be bankrupt, which is the price necessary for competition among major powers. But the price is worthless. TSMC’s production capacity is already rushed to the whole world. If you don’t buy it, you have others to buy it. It is purely for Chinese companies to sacrifice in vain. What is more likely to happen is that Chinese companies cannot find manufacturers on the mainland, so they go overseas to find them, but in the end they still give money to TSMC and spend more money.

We continue to look at the benefits of TSMC’s expansion. In addition to the most basic benefits of producing chips, there are also derived benefits, such as cultivating talents. Some people might say that Taiwanese companies guard you like thieves. How can they train you? This is a layman.

SMIC's co-CEO Liang Mengsong and vice chairman Jiang Shangyi, aren't they all from TSMC? With a bigger vision, Yin Zhiyao, the founder of China Micro Semiconductors, and Zhang Rujing, the founder of SMIC, are all from foreign semiconductor companies. With a bigger vision, how many talents in China now come from related companies abroad? Talents are cultivated through hard work. The opportunity to do hard work is definitely better than nothing. This is an obvious reason.

In addition, the expansion of TSMC's production can stimulate the domestic industrial chain. This is easy to understand, such as purchasing related equipment and raw materials. Maybe some people have to behave again, saying that TSMC will not purchase mainland China. How is it actually possible? As long as the factory is in the mainland, there are naturally ways to make it purchase.

In addition, the expansion of TSMC's production can show China's unwavering determination to expand and open up. Look at a central document, "Several Policies to Promote the High-quality Development of the Integrated Circuit Industry and Software Industry in the New Era" issued by the State Council in July 2020. The first of these is: "Integrated circuit manufacturing enterprises or projects with an integrated circuit line width less than 28 nanometers (inclusive) encouraged by the state and an operating period of more than 15 years are exempt from corporate income tax from the first to the tenth year." Look, this is an investment direction that the Chinese government clearly encourages. If you don’t even approve the projects you encourage, wouldn’t it become an international joke?

It can also be seen in this document that, let alone 28 nanometers, both 65 nanometers and 130 nanometers belong to the range encouraged by the state. This shows how backward China's chip manufacturing is. Many people think that 28nm is already outdated production capacity, and they all look down on it. Now do you understand how big the mistake is?

So the big picture is that China’s chip industry is not too strong, strong enough to pick and choose, but too weak, so weak that it needs to seize every opportunity to develop itself. More broadly speaking, we must not forget the fundamental line of reform and opening up to China that has brought great value to China because of the suppression by the United States. We must not go to the other extreme while emphasizing self-determination. We need to use every opportunity to introduce the technology and resources we need. It is always easier to let the teacher come to us than to go out and study. In terms of chips, using existing international resources as much as possible to make up for our shortcomings is an important magic weapon for rapid progress. If you can understand these truths, your knowledge level will exceed 99% of people.

After talking about these technical analysis, let's talk about the problem of thinking mode. Some people may say that TSMC is under the command of the United States, the United States is bad, and TSMC is bad. They come in to do bad things, and we must oppose it. The answer to this is that when we decide what to do and not to do, the first thing we must consider is not whether the other person is a good person or a bad person, but whether it is valuable to us. For example, during the Xi'an Incident, the Chinese Communist Party and Chiang Kai-shek had a bloody feud. Why did he let him go?

There is another wonderful example in history, how Zhuge Liang dealt with Sun Quan's proclaiming emperor. Most people may not have noticed that the Three Kingdoms period in the strict sense began in 229 AD, 6 years after Liu Bei's death. That year, Sun Quan finally proclaimed himself emperor, and before that, he was only King Wu. This posed a problem for Shu Han. If it is you, what should you do?

From the point of view, the Shu Han had 10,000 reasons for opposing Sun Quan’s proclaiming emperor: Heaven has no two days, and a country has no two kings; the first emperor Liu Bei died because of the conquest of Soochow; Liu Bei conquered Soochow to avenge Guan Yu, Guan Yu He died in a sneak attack by Soochow, and Zhang Fei was killed by his subordinates Zhang Da and Fan Qiang when he was preparing to conquer Soochow. In short, it was a bitter and deep hatred, so many people thought that they should break off friendship with Sun Quan.

But Zhuge Liang tried his best to oppose the crowd, and said something like this: "The power has a long-standing treacherous heart, so the country will be a little provocative, and ask for help. If it is more outstanding today, I will be deep... …Xian Xiaowen humbly resigned to the Xiongnu, Xiandiyou and Wumeng, all should adapt to the power and change, promote the long-term benefits, and the people who are not angry....If they are not moving and reconciling with me, my northern expedition will not be swayed by the east. The worries of Henan can’t be exhausted. This is

Profit has also been deeply rooted. The crime of arrogance is not yet clear. "

Let me translate. Zhuge Liang means: Sun Quan’s mind has long been known to our country. The reason why we don’t care is to take care of the overall situation and fight against Cao Wei together. If you tear your face with them now, they must hate us. At the beginning, the Emperor Xiandi re-allied with Soochow, it was the foresight of the mastery of power and change, and it was not like a stubborn man who was dazzled by anger. Even if Soochow didn't do anything, it was just to let us have no worries during the Northern Expedition and prevent Cao Wei from dispatching all the troops to fight against us. This benefit was already great. So for Sun Quan's crime of presumptuously proclaiming emperor, now should not be torn apart.

Zhuge Liang persuaded everyone. In the end, the Shuhan sent Wei Wei Chen Zhen to celebrate Sun Quan's claim to be emperor. The two sides should form an alliance again, and even signed an agreement to divide the world after the eradication of Wei, and then a new round of Northern Expedition began.

There are many similar stories. For example, in 1918, Lenin advocated signing the Brest Peace Treaty to cede territories to Germany in order to withdraw from the First World War and get a respite. The inspiration from these stories is that we must consider priorities. If it comes from temperament, it is the way of thinking of a child, and if you can't bear it, you will make big plans.

As a negative analogy, those who oppose TSMC’s expansion, and even label the opposite netizens as “supporting Taiwan independence”, are like Wang Ming, Bo Gu, and Li De who seem to be more revolutionary than anyone else, but they will actually cause a huge revolution. Loss. Whether you can recognize this kind of error is a big test of a person's thinking level.

Finally, we mentioned earlier that the fear of dumping and the opposition to expansion is the first level, and the second level is to recognize the situation and support expansion. In fact, there can be a higher level. If we can convince Americans that "TSMC's expansion in China is not good for China," that would be great. Let them vigorously promote TSMC to build a factory in China, thinking they can do it. This wave, this wave is in the atmosphere!