Rebirth 99 to become a car giant.

Chapter 575 Chip Empire, realized step by step
Localizing chips is easier said than done.

The so-called chip is an integrated circuit, which compresses a very complex and large number of transistor circuits into a space the size of a fingernail, thereby constructing complex circuits in a very small space and completing complex calculations and control functions.

Chips are the basis of all electronic industries and have a huge impact on human society. Whether it is as small as a mobile phone or as large as a space station, it is inseparable from the large-scale use of chips.

Chips are also related to national security. To a certain extent, it can be said that the level of chip capabilities is related to the strength of a country's weapons capabilities.

Of course, for Cao Yang, he can't think about such high-level things. What he wants is to localize the chips used in various products of Nanshan Group.

This goal is actually very difficult.

Not to mention mobile phones, there are many types of chips used in a car alone.

Especially for electric vehicles, the demand for chips is even more extensive.

Although most of the chips used in cars are not particularly technologically advanced products, chips produced with technology several years ago can also meet the requirements.

But car-grade chips do have their own difficulties.

And after becoming intelligent and connected in the future, some automotive-grade chips will also have very high requirements for computing power, and will need to use the most advanced chip technology to meet the requirements.

The construction of the Nanshan Semiconductor Magic City Fab is Cao Yang’s first step in realizing true chip localization.

"Mr. Cao, there are more people attending the commissioning ceremony of the Modu Wafer Factory this time. A vice president from the Government Affairs Council will attend the event, and there are also many people from the ministries and commissions."

"In the Magic City, relevant leaders from the city to the district are all present."

Zhang Jing has been staying in Magic City during this period. Today he came to the airport to pick up Cao Yang and report on the next arrangements.

Compared with the previous mass production ceremony at the Jaguar Land Rover Magic City Factory, this event is obviously of a much higher level.

Wafer fab!

This is not something that ordinary companies dare to do, nor is it something that ordinary companies can do.

But the Nanshan Group dared to do it, and it did it in a spectacular way!
"During the previous planning, we said that after our wafer fab is put into operation, it is best to produce 8-inch wafers first. What is the estimated production capacity of these wafers now?"

Silicon wafers come in different sizes. They used to be 4 inches, then 6 inches, 8 inches, and now the internationally advanced one is 12 inches.

Generally speaking, the more advanced the chip, the larger the silicon wafer used.

Because silicon wafers are round and chips are square, the leftover materials need to be cut off and wasted. The larger the area of ​​the silicon wafer, the less is wasted, so the cost is lower.

At the same time, the larger the silicon wafer, the more chips can be cut from one silicon wafer, and the higher the processing efficiency, which means the cost is reduced.

Chips of 28nm and below are basically all produced using 12-inch wafers. Only some relatively mature chips, such as 40nm, 65nm, etc., use 8-inch wafers.

Nanshan Semiconductor purchased the most advanced equipment in the world at that time. This was also a blessing from the financial crisis. In theory, it could produce 12-inch wafers.

However, considering that this is the first production, and the main goal of Nanshan Semiconductor in the early stage is to use various chips on automotive-grade chips, 8-inch wafers, 28nm or even 40nm technology can meet the requirements.

So in the end, Zhang Jing and Cao Yang discussed and finally produced 8-inch wafers at the Modu Wafer Factory to initially solve the Nanshan Group's needs for its own chips.

At the same time, Nanshan Semiconductor and Nanshan Equipment are also working together to localize the production equipment of the 8-inch wafer fab.

Cao Yang himself has led people from Nanshan Equipment to overcome major bottleneck equipment issues many times.

Although it is unlikely that 100% of the equipment will be localized immediately when building the second wafer fab.

However, according to Cao Yang's plan, Nanshan Semiconductor's wafer fab must have the ability to use 100% domestically produced equipment within five years at the latest.

Some domestically produced equipment may not be the best, but they can solve the problem of stuck necks at critical moments.

this is very important.

Especially for a company like Nanshan Group, which has mastered core technologies in many fields and has now begun to impose sanctions on its subsidiaries, it is necessary to consider risks in advance.

Otherwise, if some raw materials or equipment are stuck at random, your fab will be useless.

That is absolutely unacceptable.

Never overestimate the bottom line of some countries.

"It has just been put into mass production now. It is expected to take more than a month to slowly ramp up. Starting in March, it will be able to achieve a monthly production capacity of 1 3-inch wafers."

"However, according to the current demand of the automobile market, one fab cannot meet the demand. Within this year, the company can officially consider the construction plan of a second fab."

"With the experience of the first one, we can consider building another 10-inch wafer fab with a monthly production capacity of 8 wafers, or we can consider directly building a 5-inch wafer fab with a monthly production capacity of 12 wafers."

"Considering that it is relatively easy to purchase photolithography machines now, I suggest starting the 12-inch project first and purchasing a batch of the most advanced photolithography machines from ASML and other manufacturers as soon as possible."

"The equipment we are developing at Nanshan Equipment can be used in the next 8-inch wafer fab project."

Although Zhang Jing is also aware of some of the research and development of Nanshan Equipment, his energy is mainly on Nanshan Semiconductor after all. He does not think that Nanshan Equipment can handle related equipment so quickly.

He even felt that Cao Yang's requirement for chip equipment to be 100% localized was very difficult to achieve.

Of course, he would never say these words directly.

After all, deep down, he also hopes that Nanshan Equipment can handle these technologies.

"Then consider building a 12-inch wafer fab next to the Modu wafer fab, and at the same time find a place in Lingnan Province to build an 10-inch wafer fab with a monthly production capacity of 8 pieces."

Regarding the development of semiconductors, Cao Yang has already made relatively mature considerations and layouts before.

Therefore, he quickly made a judgment on Zhang Jing's proposal.

Nanshan Group is not short of money now, and it makes little sense to put too much money on the books.

Now is not the time to compete with everyone about who has the largest cash reserves on the books.

Anyway, according to the current pace, Nanshan Group's profitability in 2010 will not be better than that in 2009.

If you spend a lot of money, the funds on your account may not decrease in the end, but may actually increase.

Therefore, Cao Yang made the bold decision to launch two wafer fabs without any hesitation.

The construction of the Magic City wafer fab cost almost 60 billion.

If the monthly production capacity is doubled, the investment is estimated to be almost 100 billion.

The investment in a 12-inch wafer fab is estimated to be RMB 90 billion.

Considering the construction of these wafer fabs, Nanshan Semiconductor will definitely continue to recruit more high-end talents back. It is normal to eventually spend 200 billion in funds this year.

If you take into account the fact that the 8-inch wafer fab now in production is likely to sell wafers at a loss, it is not surprising that Nanshan Semiconductor spent 300 billion yuan this year.

It’s no wonder that ordinary companies don’t dare to easily touch projects like wafer fabs.

Even many chip companies are unwilling to invest and manage fabs themselves.

It is true that the financial and technical requirements in this field are too high, and you will lose a lot of money if you are not careful.

"Building two fabs at the same time?"

Zhang Jing was surprised and happy when he heard what Cao Yang said.

The Modu Wafer Factory has spent so much money, and it still doesn’t see any signs of making money.

He thought that Cao Yang would become cautious about the construction of the wafer fab, but he did not expect that Cao Yang would still be so bold.

This must be a good thing for Zhang Jing.

"It's possible, but equipment procurement may not be as cheap as the current factory."

"As the economy slowly recovers, TSMC and Samsung have also begun to take some actions."

"It is estimated that we will no longer be able to directly purchase lithography equipment from ASML that has been canceled by other manufacturers."

Since last year, ASML's influence in the field of lithography machines has surpassed that of Japanese manufacturers, and this trend is changing very quickly.

All chip manufacturers basically turn to ASML to buy new lithography machines.

This is because the price of photolithography machines has increased and the delivery cycle has begun to become longer.

"We can arrange for people to discuss with equipment manufacturers about purchasing equipment for 12-inch wafer fabs. At the same time, we also have Nanshan Equipment prepare for the production of equipment for 8-inch wafer fabs."

"Of course, to be conservative, we can purchase equipment from these two fabs from ASML at the same time. By then we will definitely need to continue to build more fabs to expand our chip production capacity."

"Even if you really buy it, you can still use it."

At this time, the chip equipment can be sold even if it is bought back and resold.

So there is no need to worry about whether Nanshan equipment can produce or waste money on purchasing equipment.

The main reason is that although Cao Yang has confidence in Nanshan's equipment, he is worried that he has messed up when not a single fab has officially adopted it.

After all, in the field of chip equipment, China is really not good at it.

There are a lot of holes that need to be filled in.

"No problem. After tomorrow's mass production ceremony is over, I will immediately start formal preparations for the construction of a new factory within the company."

"In the past year or so, we have poached a lot of chip talents from Japan, Korea and other places. After a period of running-in, it is time for them to start to play their role."

Zhang Jing couldn't help but rub his hands together, feeling very excited.Nanshan Semiconductor has purchased a community in both Shanghai and Yangcheng to house chip experts recruited overseas.

It also recruited a group of professionals from China and recruited a large number of fresh graduates from various universities to join Nanshan Semiconductor in order to form its own chip production capabilities as soon as possible.

This year, Zhang Jing has been busy with Nanshan Semiconductor.

"By the way, I heard that Intel and TSMC are already studying 18-inch wafers. Do you think we need to follow up?"

Although Nanshan Semiconductor has not yet built a 12-inch wafer fab, and only the 8-inch wafer fab is considered relatively backward in the industry, it does not mean that Cao Yang has no ideas.

Of course, knowing that even in 2023 12-inch wafer fabs will dominate and 18-inch wafer fabs will not be built, Cao Yang will definitely not let the people below start building 18-inch wafer fabs in a hot head. .

If it really happens like that, let alone 300 billion yuan, even if it is doubled, it will not be wasteful enough.

After all, when fabs move from 8 inches to 12 inches, there are already many successful cases in the industry that you can refer to. You can just follow them and make some improvements.

But it's different when it comes to building an 18-inch wafer fab.

The larger the wafer size, the higher the requirements for microelectronics processes, equipment, and materials.

Because about 75% of silicon wafers are produced using the Czochralski method, the larger the diameter, the greater the possibility of defects in the lattice structure due to unstable rotation speed during the crystallization process.

At the same time, the larger the diameter of the wafer, the greater the weight and the greater the risk of warping at the edges.

Therefore, the larger the wafer, the lower the yield and the higher the cost per unit area of ​​the wafer.

As a result, when the cost reduction by increasing the wafer size cannot make up for the increased cost of wafer defective rates caused by the large diameter, it becomes uneconomical to use larger-size wafers for production.

In this case, it is better to use 12-inch wafers and then devote manpower and material resources to the chip manufacturing process.

Intel, for example, is now able to produce chips using the 22nm process. Although it has not yet been mass-produced, at least it has the ability to do so.

And it is said that they have already started working on the 14nm process.

But for Nanshan Semiconductor, it now has the mass production capacity of 90nm. 65nm will not be completed until next year, and 40nm will have to wait until 2012, or even later.

Although this can already meet the needs of most car-grade chips, it is not enough for the production of main chips for Honor mobile phones.

According to Cao Yang's plan, by 2015, at least 12-inch, 28nm process chip technology must be available, and all equipment and raw materials must be localized.

After that, further expansion to the 14nm process will basically be able to meet most needs.

As for the 7nm process, there is no need to worry. Nanshan Semiconductor will have time to work on it later.

The international environment is still relatively friendly now, and chips for Honor mobile phones can still be purchased normally.

After mastering the 28nm process, we can start to make Honor mobile phones completely products that use 100% domestically produced components.

"Mr. Cao, the threshold for 18-inch wafer fabs is very high."

"On the one hand, to evolve from 12 inches to 18 inches, basically all equipment needs to be re-developed. Currently, relevant equipment manufacturers in the world have not yet made a complete breakthrough."

"On the other hand, the investment in 18-inch wafer fabs is expected to be very huge."

"Today, investing in a wafer fab can cost less than 100 billion yuan, but if it is an 18-inch wafer fab, it is expected to require tens of billions of yuan."

"The most important thing is that the purpose of increasing the wafer size is essentially to reduce costs, but the investment from 12 inches to 18 inches is likely to be more exaggerated than the cost reduction effect."

"So I think the company does not need to think about the 18-inch wafer fab for the time being. As long as the 12-inch wafer can be successfully produced in the next step, it will then focus on improving the chip production process."

Zhang Jing's answer made Cao Yang very satisfied.

It is very important that the person in charge of Nanshan Semiconductor chosen by himself does not get confused by the upcoming small victory.

Chips are sometimes not a cost issue at all for Nanshan Group, but a bottleneck issue.

If Nanshan Semiconductor can produce 8nm process chips in an 14-inch wafer factory, it will be able to meet the demand for the next 10 years.

Not to mention that it can be produced based on 12-inch wafers in the future.

Investing a lot of manpower and material resources into process improvement allows Nanshan Semiconductor to quickly catch up with the international mainstream level. This is the main mission of Nanshan Semiconductor.

As long as this can be achieved, both Cao Yang and the relevant departments of Huaxia, as well as the customers of Nanshan Semiconductor, will definitely be very satisfied.

After all, even SMIC, the leader in the domestic chip industry, now has the technology of the 65nm process.

Nanshan Semiconductor can move towards this level as soon as it comes up, which is already very good.

Of course, to some extent, this is also a late-mover advantage.

Taking advantage of the financial crisis, Nanshan Semiconductor has been poaching chip talents from various countries.

Due to the sluggish semiconductor industry in recent years, which has generally suffered losses, the difficulty of poaching people has also been reduced a lot.

A few years later, the situation may be completely different.

"You are right. In the next ten years, we will only build 8-inch and 12-inch wafer fabs. We will try our best to increase the yield rate of these two wafers and reduce the cost."

When Cao Yang said this, Zhang Jing also breathed a sigh of relief.

When the two people chatted, it was not only Cao Yang who was observing Zhang Jing, but also Zhang Jing who was observing Cao Yang.

If Cao Yang insists that Nanshan Semiconductor go in the most powerful direction, then Zhang Jing will definitely have a big headache.

After all, there are some things that you can't just pat your head on.

Even if real money is spent, it may not be possible.

In particular, Nanshan Semiconductor's first wafer fab will officially start mass production tomorrow, and the 12-inch wafer fab has not yet officially started construction.

At this time, the goal is too ambitious, and it may not inspire people, but scare them away.

"China's automobile market is expected to exceed 2000 million vehicles in the future. A traditional fuel vehicle uses hundreds of chips, while the number of chips required for electric vehicles that began to appear last year has doubled. , in the future, whether it is a traditional fuel vehicle or an electric vehicle, the demand for chips will increase."

"This means that China's annual demand for automotive chips can reach nearly [-] billion, or even more."

"Although it is impossible for our group to occupy 100% of these chip markets, there is still hope for occupying half of the market."

"After all, self-owned brand car companies purchase most of their core parts from Nanshan Group."

"For an 8-inch diameter wafer, approximately 200 chips can be produced."

"A 12-inch wafer can produce about 500 wafers."

"In this way, if our current fab produces 5 wafers per month, it will be 60 wafers per year."

"In the end, only 1.2 million chips can be produced. This does not include various defective rates. The actual number of chips that can be shipped is estimated to be about 1 million."

"In this way, it can only meet 1% of the chip demand of China's automobile market."

“哪怕是接下来修建的月产10万片的8英寸晶圆厂和月产5万片的12英寸的晶圆厂顺利的投产了，也就只能满足不到5%的市场需求。”

"Not to mention that in addition to cars, we also have related needs for a series of products such as Honor mobile phones, Nanshan equipment, and Nanshan robots."

"At the same time, we also have OEM cooperation with some manufacturers, so there is no way to use 100% of all production capacity ourselves."

"Mr. Cao, the 8-inch and 12-inch wafer fabs can already meet the development needs of our group in the next 10 or even 15 years."

Although Cao Yang affirmed Zhang Jing's proposal just now, Zhang Jing still explained the basis of his judgment in detail.

Cao Yang, who had never settled the relevant process in such detail, was shocked after hearing Zhang Jing's words.

After all three fabs are completed, they can only meet less than 3% of the demand for automotive chips?
This is only China's demand, and it is calculated based on the scale of 2000 million vehicles. In fact, the Chinese market can reach more than 2500 million vehicles.

Moreover, Nanshan Group's parts and components are not only supplied to domestic customers, but also need to be exported.

In other words, the market faced by Nanshan Group is global, which is a huge market of 8000 to [-] million vehicles a year.

Is it the current size of Nanshan Semiconductor?
Not enough to see at all!

"What you said makes sense. After listening to what you said, I think the plan for a wafer fab with a monthly output of 5 wafers in the future is a bit inadequate."

"After all three of our wafer fabs are put into mass production, the monthly production capacity of the newly planned wafer fabs must be more than 10 wafers. It is best to be able to produce 20 wafers or even 50 wafers. monthly production capacity.”

“This way we can further reduce costs and enhance our competitiveness.”

Capturing at least 50% of China's automotive-grade chip market is Cao Yang's goal for Nanshan Semiconductor.

So, at least 50 billion chips, or even 100 billion chips, must be produced every year.

The gap in the number of wafers required is still huge.

Nanshan Semiconductor has a long way to go!
"As long as we accumulate experience from these three factories, it will be much easier to simply increase monthly production capacity."

“最晚在2012年，我们就可以开始修建月产20万片的8英寸晶圆厂，在2015年之前修建月产30万片以上的12英寸晶圆工厂。”

"By then, we will strive to reach the second-largest wafer production capacity in the mainland, and even challenge the mainland's first-ranked level."

Zhang Jing was very excited now, as if he had been given a shot of blood.

The chip empire planned in my dream is expected to be realized step by step.

(End of this chapter)