When the land was acquired, the floor prices were 25,000 yuan per square meter and 34,000 yuan per square meter. Adding several thousand yuan in construction and marketing costs, the total investment in the four projects would exceed 20 billion.

Lujiazui's prime location, when it goes on sale in 2017-2018, will be high-end and improved with large flat floors, which will be 120,000 per square meter and 160,000 per square meter.

With a total investment of more than 20 billion, it is easy to earn back more than 120 billion.

The net profit is as high as 100 billion!

It can be said that the profits of 100 billion from these four plots of land alone are enough to cover the total expenditure of a 12-inch large-scale wafer fab with a production capacity of 100,000 pieces in three months, as well as the total expenditure of research and development of the 28-nanometer process!

In recent years, prices have not risen excessively, and the currency has relatively maintained its value.

To build a large-scale 28nm wafer fab with a monthly output of 100,000 wafers, 30 billion is enough.

But in a few years, building a large-scale 10-nanometer wafer fab with a monthly output of 100,000 wafers will cost 50 billion or even more.

Similarly, the current research and development of the 28nm process requires several billion yuan.

But in a few years, the development of a 10-nanometer process will cost tens of billions.

If we develop a 7-nanometer process, the difficulty factor will skyrocket several times, and the research and development costs will also skyrocket several times.

Similarly, 7 nanometers requires multiple exposure technology, and the number of DUV lithography machines required will also increase several times.

By then, both the cost of building a wafer factory and the cost of producing 7-nanometer chips will skyrocket several times.

There is nothing we can do about this.

As for the price reduction of photolithography machines?

Don't think about such good things.

In a monopolistic industry, there is no such thing as price reduction.

In a one-sided seller's market, you have to buy and buy even if the price increases.

Unless domestic DUV lithography machines are realized and fully replaced, the lithography machines will become cabbage prices.

For example, before the breakthrough of domestically produced nuclear magnetic resonance instruments, giants such as Siemens monopolized the market and sold 32 million units.

After the domestic breakthrough, the number of nuclear magnetic resonance instruments reached 2.9 million!

Similarly, the current DUV lithography machine cost 350 million units by ASML before it was domestically produced. Once fully domestically produced and technically equivalent, it can achieve less than 100 million units in mass production!

The profits from the four plots of land in the magic city could exceed 100 billion, and Wang Yi couldn't help but feel very good.

"Dr. Wang, you have gained a lot from this trip to the Magic City. How about staying in the Magic City for a few more days?" Mr. Zhou said with a smile.

Wang Yi nodded: "Yes, I'm going to go to the ASE packaging and testing factory to see if it can be acquired."

"You want to acquire a packaging and testing factory!" Mr. Zhou instantly understood Wang Yi's thoughts.

"That's right, we have two wafer fabs in Magic City, so naturally we have to acquire a packaging and testing factory. In this way, we can provide a one-stop solution for chip mass production and chip packaging and testing."

According to Wang Yi's plan, mobile phone factories/foundries, wafer factories, and packaging and testing factories will all be divided between the north and the south.

Currently, the second phase of the Xingyi Mobile Phone J1 Super Factory in Jeju has been delivered and is undergoing trial production.

It will be fully put into production by the end of the month, when the annual production capacity will increase from the original planned 16-hour non-stop production of 50 million to 24-hour non-stop production of 75 million.

The Xingyi Factory in Imperial City has also increased its annual production capacity from 30 million units with 16-hour non-stop production to 45 million units with 24-hour non-stop production.

Starting next month, Xingyi Mobile's own annual production capacity will reach 120 million, and its monthly production capacity will reach 10 million.

Such terrifying production capacity can basically cover the sales of the Xphone 2 Kunpeng Edition.

After Double Eleven, domestic monthly sales of Xphone 2 will drop to around 10 million, or even below 10 million.

Including parallel imports dumped overseas, monthly sales are expected to be around 12 million.

After all, the Double Eleven promotion can set sales records and consume a lot of potential purchasing power.

In addition, the Xphone 2 has passed its prime sales period in the first three months, so subsequent sales will naturally decline a bit.

In this way, Xingyi Technology's own factory is enough to produce 10 million Kunpeng version of xphone 2.

Add in the 2 million units of Qualcomm version of the Xphone 2 manufactured by BYD, and the total is 12 million units of the Xphone 2 per month, which is enough to sell.

As for the Xingyi Mobile Phone J2 Super Factory in Jeju, it is still under construction and will be delivered in two phases in the middle and second half of next year.

Before the delivery of the J2 Gigafactory, the Kunpeng version of the xphone 2 was produced by the Jeju J1 Gigafactory and the Teito Factory.

The Xphone 1 and Xphone 1pro, as well as the Unbounded Thousand Yuan phone, are manufactured by BYD.

This achieves the north-south diversion.

Similarly, the fab will also be divided into north and south.

The Xingyi B1 wafer fab in the Imperial Capital mainly mass-produces Kunpeng 702.

By the end of the month, the yield rate of the B1 wafer fab is expected to reach 95%, with a monthly production capacity of more than 15 million units and the ability to produce 15 million Kunpeng 702 units.

The remaining 5% of inferior chips can be made into about 800,000 low-frequency Kunpeng 700s.

After these chips are produced, they will be sent to the Magic City packaging and testing factory for packaging and testing at low cost and high efficiency through the Beijing-Shanghai high-speed rail.

After the sealing and testing is completed, the Kunpeng 702 chip is sent to the Jeju J1 super mobile phone factory, the tablet factory, and the mobile phone factory in the imperial capital via the Beijing-Shanghai line for the production of the Xphone 2 Kunpeng version and Xingyi Tablet X1 Ultra version.

The M1 wafer fab in Shanghai mainly produces low-frequency Kunpeng 700 chips and can produce more than 15 million Kunpeng 700 chips every month.

Kunpeng 700 is a castrated version of Kunpeng 702. During the R&D and design stage, the main frequency of the quad-core CPU was reduced from 1.7G to 1.3G, and the GPU was directly designed as a three-core.

Therefore, the production difficulty of Kunpeng 700 is much lower than that of Kunpeng 702, and the yield rate is naturally higher. As the proficiency level increases, it is expected to exceed 97%.

As for the Kunpeng 500, there is not even any research and development. It is just the Kunpeng 700 that uses an algorithm to lock the core and reduce the frequency, and just changed its name.

To put it bluntly, Kunpeng 500 and Kunpeng 700 are the same chip, just with the algorithm castrated.

Of course, 3% of Kunpeng 500s are really defective and are made from inferior chips of Kunpeng 700s. After all, they cannot be wasted.

More than 97% of the remaining Kunpeng 500s are actually Kunpeng 700s, which only lock the core through algorithmic core reduction and frequency reduction.

The Kunpeng 700 produced by the M1 wafer fab will be sent to the BYD foundry after being sealed and tested in Shanghai for production of xphone 1, xphone 1pro, Xingyi Tablet X1, and X1pro.

On the Xphone 1 and Xingyi Tablet X1, the algorithm will be used to lock the core and reduce the frequency, and will be renamed Kunpeng 500.

Of course, if you want to be less emasculated, you can also change it to Kunpeng 506: a quad-core CPU with a main frequency of 1G and a slightly down-clocked tri-core GPU, which is used in xphone 1 and Xingyi Tablet X1.

Next year's Unbounded mobile phone will be a little more stripped down and changed to Kunpeng 500: quad-core CPU 1G clocked, dual-core GPU, used in the 999 Unbounded mobile phone.

In this way, the Unbounded mobile phone starting at 999 and the Xphone 1 priced at more than 2,000 yuan can be distinguished, and the Xphone 1 will not be scolded for using the same chip as the thousand yuan phone.

Compared with Kunpeng 700, the performance of Kunpeng 506 has both CPU and GPU downclocking algorithms, which will not affect the sales of xphone 1pro and Xingyi Tablet X1 pro.

Anyway, they all use Kunpeng 700 chips, and it is easy to operate after algorithm castration + name change.

As for cracking the castration algorithm and achieving full performance?

The XOS system has changed the underlying architecture of Android and recoded it. If it can be cracked, it will be considered powerful.

But there is a high probability that it will become a brick!

If such a talent really existed, Wang Yi would hire him with a high salary!

More importantly, the algorithm tuning and heat dissipation of the SOC chip are based on the core and frequency of the SOC chip.

Kunpeng 500 already generates low heat through core locking and frequency reduction, so the heat dissipation module of the Unbounded phone can be adjusted casually.

Once cracked, the performance of the Kunpeng 700 is fully released, and the heat dissipation module is so severe that the Unbounded phone's heat dissipation module cannot handle it. When the phone overheats and triggers the lock-core downclocking mechanism, the performance will still be the same as that of the Kunpeng 500.

It's purely unnecessary.

The same is true for the Kunpeng 506. The heat dissipation design of the Xphone 1 has no problem running the performance of the Kunpeng 506, but if it is used on the Kunpeng 700, it is completely insufficient.

After all, the Xphone 1 is a dual-core phone, and the original cooling module was also designed for Qualcomm's dual-core. The downclocked quad-core Kunpeng 506 can still handle it, but the Kunpeng 700 with full performance cannot.

In this way, everything flows north and south.

The new flagship is produced in-house and others are OEMed.

When the sealing and testing factory in Imperial City is put into operation next year, the flagship chips produced by Imperial City will be directly packaged and tested in Imperial City, and then directly shipped to the factories in Imperial City and Jeju.

By then, everything from chip factories, to packaging and testing factories, to mobile phone factories will be divided into north and south!

This is Wang Yi's overall plan.

I had lunch with Mr. Zhou. In the afternoon, Wang Yi returned to the hotel where he was staying and held a video conference with Song Sining.

"Looking at your state, it seems that this trip to the Magic City has yielded a lot." Song Sining said with a smile.

Wang Yi nodded: "Of the nine projects, eight have been finalized!"

"???"

"Nine projects?" Song Sining was confused:

"The acquisition of SMIC's Magic City wafer fab, the acquisition of ASE's packaging and testing plant, plus the Xingyi Automobile project will support three, how can there be nine?"

Wang Yixiao said: "The acquisition of SMIC's wafer fab has been completed, but the monthly production capacity is only 25,000 wafers. The city has built us a large-scale wafer fab with a monthly production capacity of 100,000 wafers."

Song Sining: "..."

"In addition, a 300-acre industrial park has been planned in Lujiazui to serve as the headquarters of Xingyi Technology Magic City."

Song Sining: "..."
To be continued...