A 0.25 micron production line requires a high price of US$700 million. It is unrealistic and unrealistic for individuals to pay for this.
There are multiple capitals behind any chip foundry. Even China's Shougang Group claims to invest tens of billions of US dollars to build a chip manufacturing plant in Yanjing. The initial investment is US$1.335 billion.
Of the US$1.335 billion, US$120 million was invested by Shougang Group companies, and two Americen semiconductor companies each contributed US$100 million, namely AOS Semiconductor and BVI DEBORAH Semiconductor. Another 45 million US dollars was invested by Yanjing State-owned Assets Management Company, and the remaining money was obtained from bank loans.
This is a problem of risk control and profit distribution. Only by distributing profits to other companies can you obtain orders from other companies.
Zhou Xin could afford the money entirely on his own, but it would cause invisible difficulties in subsequent operations, and others would not ask you to do the work.
There are so many chip foundry companies now. Why not choose a company that I have invested in, but a sole proprietorship like Xinxin Technology? Zhou Xin has a natural halo in the Internet field, but does not have such a halo in the chip field.
And even if there is a halo, the halo will not let other companies give him chip foundry orders. No one thinks about issues based on emotions when it comes to interests.
Why TSMC can get orders from Intel is because more than 80% of TSMC's shares are held by Wall Street, and Citigroup holds more than 20% of TSMC's shares. This has also led to the fact that TSMC’s decision-makers are almost all Wanwan, and the board of directors seats are basically all Wanwan, but they are unable to refuse the request from Amerikan.
Hu Zhengming, like Zhou Xin, initially placed his hopes on SMIC. Later, he gave up the idea after carefully understanding SMIC's equity structure and the specific internal management situation.
Instead of cooperating with SMIC, it is better to build a factory by ourselves. New Core Technology will provide a small amount of money, and other companies will take the bulk. According to his plan, New Core Technology will provide 20% of the investment, and US$2 billion can leverage an investment scale of US$10 billion.
As for Chinese companies not being able to obtain the latest production technology, this is not a problem, because if they were targeting the Chinese market rather than the foreign market from the beginning, then backward production lines would not only be not a problem, but an advantage.
Hua Hong Semiconductor, SMIC and Shougang Semiconductor will only consider 8-inch, 0.25 micron production lines when building factories. They also want to import 12-inch, which is 0.18 micron production lines, but this production line has been banned.
For Americen, they do not want to cultivate another competitor in the chip field.
Americen's internal interests are definitely not monolithic. Chip equipment manufacturing companies do not mind if the competition in chip foundries becomes more intense, as their equipment is easy to sell. However, based on Americen's overall strategic considerations, they do not want China to become an important player in the chip foundry field.
Starting from guided missiles, Washington strategists have realized that chips and military capabilities are closely related, even highly correlated. Therefore, from a commercial perspective, China can obtain chip production equipment, but from a strategic perspective, these equipment and technologies must be at least one generation behind.
Of course, Neon and Americen have different ideas. They have already lost their original status in the chip manufacturing field. In the 1980s, the market shares of NEC, Toshiba, Hitachi, Fujitsu and Panasonic, which were all-powerful, were continuously squeezed by Samsung and TSMC.
Neon is retreating from manufacturing in the chip field and becoming a provider of raw materials and equipment, so they don't mind cooperating with China. NEC even hopes to make a comeback through China's vast market.
So instead, neon manufacturers hope to break away from the constraints of the Wassenaar Agreement.
In the 1980s, Neon's chip industry accounted for half of the entire chip industry and 45% of the world's semiconductor market. Until the early 1990s, 6 of the top ten semiconductor companies were still from Neon. In 1994, Sony's global shipments of color TVs reached 100 million units. What an astonishing number.
Matrix is still a long way from Sony's pinnacle position in consumer electronics. In the 1980s, Neon's memory chip, also known as DRAM, defeated the American company, and more than 80% of the American DRAM companies went bankrupt.
In 1980, the Neon Electronic Machinery Industry Association held a seminar in Washington to promote DRAM produced by Neon. Representatives from Hewlett-Packard greatly praised the DRAM produced by Neon at the meeting. In the semiconductor industry of Amerikan It caused a strong shock. So much so that the editor of Amerikan's technical magazine "Electronics" was incompetent and furious:
"Neon Semiconductor Company came to teach American Company how to conduct quality management, and American Company representatives personally proved that what they taught was correct!"
Of course, it is another story that Amerikan later interfered in Neon's semiconductor business operations on the grounds of national security.
Hu Zhengming is almost familiar with this period of history. He knows many frustrated neon engineers at the annual conference held by IEEE. They often recall the neon semiconductors in the 1980s that were so exciting, but reality has dragged them down. Back to alcohol.
Therefore, Hu Zhengming does not dislike the 8-inch production line. In his opinion, the 8-inch production line can already do a lot of things.
It is true that the 8-inch production line produces 0.25 micron chips, but by optimizing the process, structure and function, it can theoretically achieve 0.13 micron, which is even lower than the 0.18 micron of the 12-inch chip.
As for why we insist on choosing Huahong, it is because Huahong’s 8-inch production line is still in the hands of Huahong itself, not Huahong NEC. Huahong is considering partners.
Xinxin Technology is of course a good partner. No matter from which aspect, Zhou Xin behind Xinxin Technology is an excellent partner. He controls Matrix, a consumer electronics company that shows great potential. It has a strong influence in Silicon Valley, has an extremely large cash flow, and has a good relationship with senior executives in China.
There is a natural basis for cooperation.
Unlike Shougang's 8-inch production line, which is still in planning and has just completed the preliminary planning work, Huahong's 8-inch production line is about to be put into use.
If there is no Xinxin Technology, Huahong's 8-inch production line will be allocated to Huahong (American) Company.
Hu Zhengming hopes to build this production line into a model project and attract more Chinese capital or international capital to join in the future.
This cooperation model can be adopted in the future, so the most important goal at the moment is to convince Hua Hong.
And the control rights, that is, the mechanism of different rights for the same shares, must be written into the contract.
Neither Hu Zhengming nor Zhou Xin wanted to limit such a crucial point of control to just verbal restraint. They allowed Huahong to do this, so will the subsequent chip production lines jointly established by Xinxin Technology and other companies be just verbal restrictions?
Jiang Shoulei said that the control of Huahong NEC is in the hands of NEC. The middle-level managers are almost all Neon people. This kind of control is only temporary. In May 2003, Huahong will take back the entrusted management rights from NEC. .
Hu Zhengming didn't know what was going to happen in the future, but based on experience, contract restraint was the safest method.
“The big reason why Xinxin Technology is obsessed with control is that we believe that there may be big differences between Huahong and Xinxin Technology in terms of future business strategies.
Under the contract constraints, we will ensure that the annual revenue is no less than 2 billion RMB, and we hope to invest these revenue in research and development. Then we hope to tilt the salary package towards front-line employees and R&D personnel.
Then we will invest the funds in some of our upstream and downstream domestic suppliers. We will not rush to pay large dividends. Of course, the dividend itself will definitely be done, but it will not be so fast.
Therefore, we do not want to have to agree with Huahong on these matters in the future. This will seriously affect the efficiency of corporate operations. Huahong can monitor where every payment goes.
Huahong has the power of supervision. Newly established companies can be audited by Huahong every year, and capital transactions can be connected to UnionPay’s capital management and control system. We can make progress in these areas, but different rights for the same shares must be written into the contract. "
After Hu Zhengming finished speaking, Jiang Shoulei was completely willing to cooperate with Xinxin Technology from the perspective of a semiconductor practitioner.
Under the premise that the entire semiconductor industry is in recession, no company except New Core Technology can write guaranteed revenue into contract terms.
Although there is a time limit, the time limit for revenue guarantee in the contract is five years, 2 billion yuan a year, and the gross profit margin is guaranteed to be no less than 30%. In other words, Xinxin Technology cannot use a low-price strategy to ensure revenue.
Five years is long enough. You must know that chip production lines require equipment depreciation every year. For a chip production line purchased for US$700 million, the entire value of the equipment will be depreciated in about five to seven years. In other words, in terms of company assets, the book value of this production line, which was actually purchased for US$700 million, will return to zero.
Amortize its cost into the previous annual cost. If the joint venture between Xinxin Technology and Huahong completes the depreciation of the production line in five years, then the cost of subsequent chips produced by this production line will only be labor and Materials cost.
Among chip foundry companies, the bulk of the cost is equipment depreciation. This is why TSMC's profit margin is as high as 60%, while the profit margin of other domestic chip manufacturers is only 30% at most. This is because TSMC's mid- to low-end production lines have completed the depreciation of production equipment, and there is no longer the cost of depreciating the value of the equipment.
Most of the production equipment of other domestic chip manufacturers began to come online in 2019.
Moreover, the bulk of TSMC's revenue comes from mid-to-low-end chip foundry, while high-end chip foundry only accounts for a small part. Even in the 3nm era, almost only mobile phone manufacturers can afford TSMC's 3nm chips.
So no matter which angle you look at, Xinxin Technology’s conditions are too superior.
Jiang Shoulei sighed: "Professor Hu, I personally am very willing to cooperate with Xinxin Technology, and I will try my best to promote this matter.
But this is not something I can have the final say on. This is more of a political issue than an economic issue.
On the economic level, to be honest, China's investment in the chip field on a national basis has resulted in considerable losses.
The conditions given by Xinxin Technology have been determined. Huahong's 8-inch production line will at least not suffer losses. In the current environment of the chip industry, it is very rare to be able to avoid losses.
This is a brand new way. It is a joint venture with a private enterprise or a method of same shares but different rights. I am not sure whether the higher authorities will allow Huahong to become a pilot. "
Hu Zhengming reacted quickly: "Mr. Jiang, you can regard it as Huahong's venture investment in Xinxin Technology.
It's just that in this venture capital investment, Huahong invested not money, but a production line, but it was still venture capital in essence. The revenue and gross profit margin guaranteed by Xinxin Technology are the agreements reached before Xinxin Technology accepts this venture capital investment.
Do you know NewPay? "
Jiang Shoulei nodded: "I know, NewPay also has some business in China, and they are very active in the foreign trade industry in the Yangtze River Delta region. I know some professors from Fudan School of Economics, and they speak very highly of NewPay and think that NewPay has created a brand new economic model.”
Hu Zhengming continued: “Yes, NewPay has accepted investments from Citigroup, Wells Fargo, JPMorgan Chase, Goldman Sachs and other institutions, and the investments from these institutions all adopt a structure of different rights for the same shares.
The voting rights between the equity held by them and the equity held by Zhou Xin are 1:10.
Therefore, if this is regarded as venture capital, there is nothing special. China has ambitions in the chip field and hopes to occupy a place in the chip field. However, the high investment from the government level has not achieved the desired results. I think it’s time to switch to another way of investing.
Treating the cooperation between Xinxin and Huahong as a pilot is exactly in line with the spirit of Shenhai and Zhangjiang in the field of innovation. "
The same-share-different-rights model is new in Silicon Valley, because Google is far from as successful as it was in the past, and even needed Wall Street to push them to merge with Baidu to gain their original status. Therefore, until 2001, the A/B share structure was not popular in Silicon Valley.
When entrepreneurs propose that they wish to adopt an A/B share structure, investment institutions will ask: Are you Newman?
When China really began to have the concept of different rights for the same shares, it should be 2013. At that time, the "Guiding Opinions on Carrying out the Pilot of Preferred Shares" was issued. Compared with ordinary shares, preferred shares have priority in the distribution of profits, priority in the distribution of residual properties, etc. Advantage, in fact, this is a kind of different rights for the same shares.
The domestic capital market will not accept the A/B share structure until the Science and Technology Innovation Board listing rules are promulgated, which clearly stipulates in Article 2.1.4 that different shares can have different voting rights.
And the reason why Jiang Shoulei is very excited about cooperating with Xinxin is that he wants to promote this matter. With the 8-inch production line in Huahong's hands, it is far from possible to create annual revenue of 2 billion RMB, let alone maintain this amount every year for five years.
He dared not and could not agree because in 1995, the state-owned assets management department clearly stipulated: "It is not allowed to agree or approve plans for state shares, state-owned legal person shares and individual shares to have different rights and benefits for the same shares."
After Jiang Shoulei and Hu Zhengming finished talking, Hu Zhengming did not limit the matter to his own level. He knew that when it came to model innovation, a higher-level driving force was needed, and Zhou Xin needed to make an effort at this time.
Zhou Xin was not polite. He first made an appointment with Director Zeng via email, and then made a trip to Yanjing specifically for this matter to have an interview with Director Zeng.
After Zhou Xin explained in detail, the other party said: "I think we can use Huahong and Xinxin Technology as pilot projects. The system is dead and people are alive. Since neither the 908 Project nor the 909 Project has achieved the expected results before, Effect.
It is even far from our expectations. Not to mention cultivating a number of chip companies, even the companies we invested heavily in establishing are having such a hard time surviving. It is time to make changes. The conditions offered by Xinxin Technology are also very favorable.
I agree with this in principle and will even vigorously push for its approval.
But I advocate that there should be process compliance, because you just told me, Zhou Xin, that we are exploring model innovation. Since it is model innovation, this cooperation between Xinxin and Huahong will provide a good foundation for subsequent state-owned enterprises. Establish samples of cooperation with private enterprises.
Then we need to follow the template starting from the prior process.
I think the following principles need to be followed: Full demonstration must be conducted in advance. Since the same shares with different rights are to be operated, a precise investment demonstration must be carried out. There must be sufficient evidence to prove that in this project, the same shares with different rights will be better than the same shares with different rights. Rights are more valuable.
This value is mainly reflected in two aspects: first, the value of obtaining project opportunities. Simply put, if the same shares have different rights, it is impossible to obtain this investment opportunity, and naturally there will be no investment income and strategic value; It means that the economic benefits obtained from the same stock with different rights can meet the normal return level of general investment behavior.
The second is to make decisions in compliance. Strictly follow the investment decision-making process of state-owned enterprises, and keep various decision-making materials and decision-making meeting minutes during the process so that they can withstand the test of audit;
The third is to insist on openness and transparency. Different rights for the same shares is not prohibited by law, so there is no need to hide it. It should be boldly written into investment agreements, project cooperation agreements, company articles of association and other documents, and do not make any drawer agreements.
The fourth is to insist on fairness and justice. Whether a state-owned enterprise gives up part of its voting rights or dividend rights, or a partner gives up, the principle of consistency of rights and responsibilities must be adhered to, that is, the shareholder liability corresponding to giving up part of the rights should also be exempted accordingly.
This is my personal initial idea, I will push this as soon as possible, because the places I can think of are only a small part, and there will definitely be missing things.
We will formulate relevant notices and guidance as soon as possible, and then the cooperation between Huahong and Xinxin Technology can first follow the guidance.
We will continue to pay attention to this case in the future, and then adjust the policies we will eventually introduce to truly innovate the model. "
Model innovation is very important. In my opinion, our current way of investing in large chip funds can only achieve catch-up, but it is difficult to achieve innovation.
And then there will be more tomorrow morning
(End of chapter)
