Rebirth 08: Rise from copycat phones

Chapter 636 The 2.5 Trillion Yuan Semiconductor Plan
As September began, Zhuyun Group officially announced a new round of major investment projects, including several previously announced major investment plans such as computing center infrastructure construction, advanced semiconductor factories, and investment plans within advanced packaging factories.

The investment plan is expected to exceed 2.5 trillion yuan, an additional 500 billion yuan compared to the previously announced 2 trillion yuan plan.

This is also an obligation of listed companies: to disclose major events in a timely manner for investors' reference!
Otherwise, if listed companies make reckless and haphazard investments without disclosing relevant information, it will mislead investors.

As a dual-listed company, Zhuyun Group will also announce its investment plans at an appropriate time.

Of course, only the relevant surface information will be released, and the detailed plan will not be disclosed... The detailed plan involves trade secrets.

In this publicly disclosed investment plan, Zhuyun Group revealed for the first time its investment plan for Zhuyun Microelectronics' 41st factory.

The investment in a single factory is expected to exceed 220 billion yuan. Despite the fact that this factory uses double exposure and has relatively low production efficiency, the monthly production capacity is still planned to reach 50,000 pieces per month.

This 41st factory is just one of the investment projects of Zhiyun Group in advanced semiconductor technology.

In this major investment plan announced in early September, Zhuyun Group plans to invest 2.5 trillion yuan over the next three years to build advanced semiconductor production capacity and conduct corresponding technology research and development.

The main investment projects are investments in the equivalent three-nanometer process. In addition to the newly built Plant 41, there is also Plant 43. These two plants are expected to use the second-generation N3 process. In addition, there are upgraded and transformed Plants 32 and 29, which will use the first-generation N3 process.

The headquarters of Zhiyun Group has set a production capacity requirement for Zhiyun Microelectronics: to increase the equivalent 3-nanometer process capacity to more than 150,000 wafers per month within the next three years to meet the production needs of mobile phone SOCs, CPUs and various computing chips.

The plan also announced the official launch of the equivalent 2-nanometer project, with the first factory to begin construction by the end of this year. This is a next-generation factory planned to use EUV lithography machines with new technology, and the expected investment scale of the first 2-nanometer factory will reach US$50 billion.

In addition to the massive investment plans for 3nm and 2nm factories, there are also investment plans for capacity expansion and upgrading of existing advanced processes... namely, the two process nodes of equivalent 5nm and equivalent 7nm processes.

These two process nodes are still the most advanced technologies at present and are the main production processes for various major chips under the Zhuyun Group.

Therefore, not only is its own demand large, but the demand for external OEM manufacturing is also increasing...

A large number of chip design companies are gradually moving into the 7-nanometer and even 5-nanometer process market... The top think tanks of the 'Zhiyun Strategic Economics Department', the global strategic economic analysis institution under the Zhiyun Group headquarters, submitted a semiconductor industry analysis report to the group headquarters after tracking and analyzing global economic development and industrial technology development.

This analysis report points out that 7-nanometer and 5-nanometer processes will be the mainstream processes in the next ten years or even longer, with a huge market size.

This analysis report, along with the business prospects plans of various business units within the group and the future market forecast analysis from Zhuyun Microelectronics, ultimately led Zhuyun Group to decide to continue expanding its 5nm and 7nm process capacity and to continuously promote technological upgrades, especially technological upgrades.

Zhiyun Microelectronics has been continuously upgrading its existing processes. The 7-nanometer process has evolved to the third generation, and the 5-nanometer process has also evolved into the second generation (N4).

There are even more variations of the 14nm process node, including the early N14, then N12, and then the N12LP process which focuses on ultra-low power consumption, and the N12S process which focuses on high performance.

The 28nm process node includes N28 process, N28LP process, and N25 process.

However, the 32nm, 18nm, and 10nm processes have not undergone further technological evolution because these process nodes are all transitional processes. Zhuyun Microelectronics has already stopped producing the 18nm process. Although the 10nm process is still in production, the scale is not large, and its production lines have been gradually converted to the N12LP and N12S processes.

The reason for continuously expanding production capacity and upgrading technology for existing 7nm and 5nm processes is twofold: firstly, the demand for chips is large, and secondly, the market demand for external foundries is gradually increasing!

As the world's largest and most technologically advanced semiconductor manufacturer, Zhiyun Microelectronics' massive production capacity not only meets its own needs but also satisfies the foundry needs of a large number of customers.

In addition to domestic clients, the number of orders from American chip design companies has been increasing in the past two years. Last year, ADM officially handed over the foundry order for its new generation flagship CPU to Zhiyun Microelectronics.

Tesla's chip orders have also been placed with Zhiyun Microelectronics!

Even Qualcomm's new generation of automotive-grade computing chips is actually made by Zhiyun Microelectronics... although their automotive-grade computing chips have very poor sales!

The return of many foreign customers to Zhiyun Microelectronics to have Zhiyun Microelectronics manufacture advanced chips is also closely related to the overall development of advanced semiconductors!
Zhiyun Microelectronics has a huge advantage in terms of technology, capital, and production capacity.

Meanwhile, Zhiyun Microelectronics is moving too fast ahead, while TSMC, which was once highly regarded by overseas chip design companies such as AMD, Apple, and Qualcomm, has been increasingly underperforming in the past two years...

As for Samsung, it goes without saying that their self-proclaimed 5-nanometer process technology is extremely poor, with a transistor density of only 120 million transistors per square millimeter. In comparison, Zhiyun Microelectronics' N5 process has a transistor density of 200 million transistors per square millimeter, while TSMC's has a density of over 170 million transistors per square millimeter.

The so-called 5-nanometer process of Samsung is actually only slightly better than the second-generation 7-nanometer process of Zhiyun Microelectronics in terms of performance, but the difference is limited, while the cost is much higher than that of Zhiyun Microelectronics' second-generation 7-nanometer process.

Currently, the only company that can barely keep up with Zhiyun Microelectronics in terms of technology is TSMC... but it is becoming increasingly difficult for TSMC to catch up, and the technological gap is widening.

There are many reasons for this situation. In addition to Zhiyun Microelectronics' talent advantage and faster technological advancement, it is also closely related to the changes in the overall semiconductor market and even the smart terminal market.

On the one hand, its advanced process manufacturing business has been greatly suppressed by Zhuyun Microelectronics: Zhuyun Microelectronics has too much technological advantage and also has a certain price advantage!

Most importantly, Zhiyun Microelectronics can provide large-scale advanced process capacity, especially 7nm and 5nm processes! Chip manufacturers seek other TSMC foundries to manufacture 5nm chips because TSMC's capacity is very limited, and it takes six months or even a year to schedule production.

However, with Zhiyun Microelectronics, they can usually schedule production for you in two or three months, and if you're lucky, they can even schedule production for you in one month.

When comparing price, technology, and production capacity, Zhuyun Microelectronics has a significant overall advantage in competing for third-party foundry orders, making it difficult for TSMC to compete with it!
This has suppressed the development of TSMC's foundry business... especially in the two advanced processes of 7nm and 5nm, which require huge investments but also have relatively high profits.

In addition, another very important point is that with the rapid development of Zhiyun Group, the rapid development of its various smart terminals, export semiconductors, robots, virtual devices and other businesses has not only created a huge demand for advanced process chips for itself, but has also suppressed the advanced chip market of other chip design manufacturers!
For example, the global demand for mobile phone SoCs is basically fixed, at around 1.3 to 1.4 billion units per year... Among them, Zhuyun Semiconductor, a subsidiary of Zhuyun Group, can obtain at least 70% of the market share.

Why does it have such a high market share?

A mobile SoC needs to be integrated into a mobile phone, so you can tell by looking at mobile phone sales figures...

The sales volume of S-series and A-series phones equipped with S-series chips, plus the SX/SXL series phones, is around 270 million units per year.

Then there are the W-series chips, which are used by a wide variety of manufacturers.

Our own clients include Xinyun Phone and Vcool Electronics Group, totaling approximately 260 million units.

The two companies controlled by Xu Shenxue alone consumed 520 to 530 million mobile phone SOCs.

Then there's Huawei in China. This company uses some W-series chips and some of their self-developed chips, but their self-developed chips are exclusively manufactured by Zhiyun Microelectronics.

Next are OV and Xiaomi. Eighty percent of the mobile phone SoCs from these two companies are purchased from Zhiyun Semiconductor's W-series chips, and about twenty percent are from Qualcomm chips.

Transsion, which mainly targets the low-end overseas market, can sell tens of millions of smartphones a year. They all use W-series chips from Zhuyun Semiconductor.

Then there are companies like Lianxiang (Motorola), Sixing, and LG. These companies' mobile phone SOCs, and Zhiyun Semiconductor's W-series chips, can probably occupy a market share of 20% to 40%.

In addition to other second-, third-, and fourth-tier mobile phone manufacturers, Zhiyun Semiconductor's W-series chips are also able to occupy a certain market share.

In conclusion, Zhuyun Semiconductor's S-series and W-series chips account for approximately 70% of the global market share.

All of these mobile phone SoCs are manufactured by Zhiyun Microelectronics... and then there's also Huawei's self-developed mobile phone SoC, and there's also a domestic low-end mobile phone SoC company whose low-end chips are all manufactured by Zhiyun Microelectronics!

This means that Zhiyun Microelectronics manufactures at least 75% of the world's mobile phone SoCs!

The global production and sales volume of mobile phone SoCs is almost fixed. The fact that Zhiyun Microelectronics manufactures so many mobile phone SoCs means that TSMC and Samsung do not have sufficient manufacturing orders.

The mobile phone SoC manufacturing market is just one typical example, and similar situations exist in many other types of chips...

Not to mention that the computing power chip group, which currently has the largest growth rate, has almost formed a de facto monopoly. These orders for computing power chips are all outsourced to Zhuyun Microelectronics, and it is impossible for them to be outsourced to TSMC or Four Star Semiconductors.

Simply put, if there were 100 in the global advanced process chip foundry market, Zhuyun Microelectronics, relying on the Zhuyun Group, would take at least 80%.

With the remaining 20% ​​market share, TSMC still has to compete with Samsung... This market share is not large, so revenue growth is limited, profits are also limited, and with less money earned, the funds available for investment in advanced processes are naturally insufficient.

This isn't just a few tens or hundreds of billions of dollars, but a massive sum of hundreds of billions or even trillions of dollars...

After all, if you're competing with Zhiyun Microelectronics for advanced semiconductor technology, and you're only investing a few billion dollars, Zhiyun Microelectronics won't even bother to consider you a competitor.

If it's a matter of several hundred billion US dollars, Zhiyun Microelectronics will pay some attention, but the level of attention will be limited... because with several hundred billion US dollars, at most one or two advanced process semiconductor factories can be built, and if there is no further follow-up, the impact will actually be limited.

To truly have a significant impact on Zhiyun Microelectronics... you'd have to invest hundreds of billions of dollars to build several large semiconductor factories, along with a bunch of supporting packaging and other types of factories!
Moreover, it requires continuous investment of money... because the technology is constantly advancing, which also means that new factories need to be built continuously!
If you don't build a new factory, Zhiyun Microelectronics will be able to completely leave you behind in two or three years... For example, Zhiyun Microelectronics has already started to play with 3-nanometer process technology. Will you follow suit?
If you follow suit, the R&D costs will start at several billion dollars, and if things don't go smoothly, they could even burn through tens of billions of dollars.

Tens of billions of dollars are poured in, and now that the technology is developed, factories are being built. You glance at the investment cost list for the three-nanometer process, and it starts at two or three hundred billion dollars.

If more production capacity is desired, investment can easily exceed 70-80 billion US dollars.

Are you going to join or not?

Well, you've barely caught up with the three-nanometer era... Before you can even catch your breath, Zhiyun Microelectronics is already announcing that its two-nanometer technology is undergoing verification.

Are you going to join or not?

At this point, it became even more expensive, with research and development costs starting at tens of billions of dollars!
The estimated investment of $30 billion for a single factory is not enough; it should be at least $40 billion or $50 billion. To increase production capacity, the investment should start at $100 billion... Are you going to follow suit?
According to Zhiyun Microelectronics' investment plan, after the equivalent two-nanometer process, they will continue to develop the equivalent 1.5-nanometer and equivalent one-nanometer processes... The subsequent investment is astronomical, and at this point, having money is not enough; the supporting semiconductor equipment and consumables industry chain also needs to keep up.

For example, current EUV lithography machines are definitely not enough; we must use more advanced lithography machines, such as the large numerical aperture EUV lithography machine, HEUV-800 series, which Gulf Technology is developing!

There are also etching machines, plasma implanters, mask machines, thin film deposition equipment, ultra-high precision cleaning machines, ultra-high precision CMP (polishing machines), and heat treatment equipment.

Then there are supporting packaging and other key equipment for processes, such as single crystal furnaces and oxidation furnaces for manufacturing silicon crystals.

The entire advanced process of chip manufacturing covers numerous steps, including front-end, back-end, packaging, testing, inspection, and analysis. It's not enough to just have one chip lithography factory; you also need a whole bunch of supporting factories with the same top-notch technology.

Behind this lies a complete supply chain of equipment for advanced semiconductor processes...

Well, we've only talked about semiconductor equipment so far; semiconductor consumables will follow later.

The development of a massive semiconductor supply chain also requires a huge influx of capital.

Backed by the behemoth Zhiyun Group, Zhiyun Microelectronics has plenty of money to invest, with hundreds of billions of dollars at its disposal. Then, through Fairy Mountain Holdings, it drives the continuous spiral development of the entire industry chain.

TSMC and other companies used to have this advantage, but they don't anymore... They're almost out of the market for advanced process chips, so where will they get the money to continue driving the development of the entire industry chain?
When technology advances further and costs reach hundreds of billions of dollars, it's no longer something you can simply follow if you want to!
No money, what's the point of following along!
And another very important point is that as the investment cost of advanced processes increases, even if TSMC catches up and has the technology and capacity for 2-nanometer and 1-nanometer processes, TSMC will not be able to find customers to pay for it!

Because by then, the cost of advanced chips will be terrifying, and conventional end products will no longer be able to afford such high prices!

In Zhiyun Semiconductor's S-series chip design plan, the chip process for several future generations will be based on the three-nanometer process... There are no plans to use the two-nanometer process in the short term, let alone the one-nanometer process, as the cost is too high and it is not feasible to use it in the short term.

Zhiyun Microelectronics is developing advanced processes like 2nm and 1nm, not for mobile phone SoCs, but for computing power chips... Currently, only high-margin computing power chips can afford the high costs of these top-tier processes in the future.

At this point, a very awkward situation arose: Zhiyun Group had almost monopolized the computing chip market!
TSMC and others can't expect Zhiyun Group to outsource the manufacturing of their computing chips to them!

This means that even if TSMC plans to invest in two-nanometer technology now, it will still face the awkward situation of having no customers.

The semiconductor industry is also a large-scale mass production industrial system. It cannot survive independently and must rely on industries that use advanced chips extensively for mutual dependence.

With its various computing chips, smart terminal and PC businesses, plus a certain export CPU industry, Zhiyun Group has ample motivation and sufficient funds to support the continuous development of advanced semiconductor technology... This is an advantage that other manufacturers do not have.

Therefore, when Zhuyun Group announced its massive semiconductor investment plan in September, the semiconductor industry fell silent... They simply couldn't keep up!
Of course, despite the reality, we cannot lose our fighting spirit!
A few days later, Intel executives revealed in a press interview that Intel was actively advancing its advanced process technology plan.

Meanwhile, TSMC officially announced its 3-nanometer process technology at a press conference, stating that it will undergo technical verification by the end of this year at the latest, and that it has already reached cooperation agreements with strategic customers... These strategic customers are naturally Apple and Qualcomm.

Looking globally, besides Zhiyun Group, only Apple and Qualcomm have a huge demand for 3nm process technology in the short term...

However, TSMC chose to avoid discussing the two-nanometer process, the 1.5-nanometer process, and the one-nanometer process that follow the three-nanometer process.

Because its core customers, Apple and Qualcomm, have made it clear to TSMC that the two-nanometer process is too expensive and they do not plan to use it for several years... because they have also learned that its biggest competitor, Zhuyun Semiconductor, does not have a two-nanometer SOC plan in the short term, but plans to stay in the three-nanometer stage for several years.

Since Zhiyun Semiconductor is not continuing its progress, Apple and Qualcomm can also take a breather and pause their 3-nanometer process as well.

Over the years, they have followed Zhuyun Semiconductor in continuously promoting the technological advancement of mobile phone SOCs, which has cost a lot of money and effort.

Now, Zhiyun Semiconductor is sending them a hint that they should slow down a bit... They certainly won't be foolish enough to unilaterally start a chip manufacturing process race: three nanometers is fine, let's slow down, make some money for a few years, and recoup our losses before talking about anything else!

Then... TSMC lost its startup customer for its future 2nm process!

Therefore, they can only develop the technology while waiting for the right opportunity... It's impossible for them to foolishly invest several hundred dollars or more in some disodium nanotechnology without having any customers.

If TSMC is in such a state, then Samsung is even more so. They can't even get the 5nm process working properly, and they don't even have a clue about the 3nm process, let alone the 2nm process.

As of September, only Zhiyun Microelectronics is continuing to advance the technology of 2-nanometer and smaller processes in the global semiconductor industry.

Besides the four top-tier semiconductor manufacturers mentioned above, there's no need to even mention the other second- and third-tier semiconductor manufacturers, who haven't even mastered the 7-nanometer process yet!
Although SMIC has tried to develop 7-nanometer process technology, it has been unable to break through the key technologies. This year, they also experimented with 7-nanometer process, but the yield was extremely low and it had no commercial value at all.

The development of 7-nanometer process technology is very difficult, and they don't have enough funds... Investing hundreds of billions of dollars is too exaggerated.

Most importantly, commercialization is difficult: We currently have Zhiyun Microelectronics, which is the world's number one and has the most advanced technology, and TSMC and Samsung also have mature and complete 7-nanometer process capacity...

Winning foundry orders for 7nm process technology from these three major semiconductor manufacturers is extremely difficult.

Therefore, SMIC's 7nm process project has actually been suspended... and it has begun to focus on the optimization and upgrading of the 14-12nm process node.

Similar situations exist with GlobalFoundries, UMC, and others. They have successively announced the suspension of 7nm process research and development, and instead focused on the optimization and upgrading of 14/12nm process nodes, and diversified their development in fields such as RF and embedded memory.

The domestic second- and third-tier semiconductor manufacturers are basically dedicated to mature processes, especially 28-45 nanometer processes. A few are more ambitious and are working on 14 nanometer processes. As for 7 nanometer processes, that's beyond their reach.

The current global semiconductor manufacturing landscape is closely related to the powerful industrial influence of the behemoth, Zhuyun Group.

The reason is extremely simple: Zhuyun Group holds a massive number of orders for advanced process chips...

However, Zhiyun Group's continued push for semiconductor development is no longer for the smart terminal industry or PC-related businesses, but for various businesses within the artificial intelligence system, including exporting APO graphics cards and automotive-grade chips, as well as supporting its own two new core businesses: intelligent robots and virtual devices.

The fact that mobile phones may not be able to afford 2-nanometer or 1-nanometer chips in a few years does not mean that intelligent robots or virtual devices will be unable to afford them, nor does it mean that various enterprises cannot afford to purchase APO series graphics cards!

Especially the APO series graphics cards!

Major companies don't care whether your unit price is high or low; they care about the computing power cost. As long as the computing power cost is more cost-effective, they will still buy APO graphics cards that cost one million or even more!
The driving force behind advanced semiconductors in the future will no longer be mobile phone SoCs, but computing chips... and artificial intelligence technology.

This is also the core reason why Zhiyun Group is willing to spend huge sums of money to invest in the construction of the 41st and 43rd second-generation 3-nanometer process factories, as well as the construction of the 42nd factory, a new generation of high-bandwidth memory, which is called 'BHM4' internally by Zhiyun Storage. At the same time, it is prepared to spend even more money to promote 2-nanometer process technology.

(End of this chapter)