Rebirth 08: Rise from copycat phones

Chapter 284: Thirty Billion in Three Years

Although the stock price fell instead of rising after Zhiyun Group released its financial report, it still failed to cause a huge sensation!
Zhiyun’s revenue and profits are really too exaggerated!

This is a revenue of more than 290 billion US dollars, ranking fourth among all domestic companies on the Fortune Global 500 list.

And what company is ahead of Zhiyun?
Two oil companies and one power grid company.

Profit ranked first, reaching 51 billion US dollars.

At the same time, according to the financial report released by Zhiyun Group, the total assets of the group are more than 285 billion US dollars, and the asset return ratio has reached 17.8%.

Compared with the single-digit return on assets ratios of a large number of traditional large enterprises, Zhiyun Group's return on assets ratio shows the high-return characteristics of a very typical high-tech enterprise.

In comparison, the share of fruit is around 20%... There's no comparison. They are a typical asset-light enterprise. It's an exaggeration to say that they are an assembly plant because they don't even have a factory.

In terms of business type, Zhiyun and Sixing are more similar, and Sixing's return on assets is around 12%.

However, Zhiyun's asset-lightness is higher and its earning ability is stronger than that of Four Stars.

Overall it's between fruit and four stars.

Zhiyun Group's core profits mainly come from the field of smart terminals. Zhiyun's smart terminals mainly take the high-end route, with high gross profit margins and huge sales volumes, which greatly support the group's overall profitability.

If the semiconductor business is taken into account, then Zhiyun's situation is much worse, with an asset return ratio of only a few points.

The main reason is that the investment in advanced semiconductor technology is too great. At the same time, Zhiyun Microelectronics has failed to gain a leading advantage. It even lagged behind TSMC and other factories in the previous 28-nanometer process, resulting in a disproportionate income and investment in chip manufacturing.

The reason why Zhiyun's return on assets ratio is not as good as that of Fruit, its profit is lower than expected, and its market value is relatively low compared to its huge revenue is due to the drag of its semiconductor manufacturing business.

However, even though the semiconductor manufacturing business requires large investments but makes little or even no money, the Zhiyun management team with Xu Shenxue as the core remains steadfast in pushing for continued investment in the semiconductor manufacturing business.

The outside world didn’t even know that shortly after Zhiyun Group announced its financial report, Xu Shenxue convened the group’s core senior management to discuss investments in the semiconductor field this year and in the next few years.

At the meeting, Ding Chengjun, CEO of Zhiyun Microelectronics Co., Ltd., introduced Zhiyun Microelectronics' subsequent major investment plans to everyone present: "This year, in addition to continuing to invest in the construction of the existing wafer fab No. 18 project, we will also start another wafer fab project, wafer fab No. 21, in the first half of this year, and a new project, wafer fab No. 22, in the second half of this year."

"In addition to the three logic chip factories mentioned above, we will also invest in the construction of a memory factory, Factory 19, and a flash memory factory, Factory 20."

"These three newly built factories, including the logic chip factories, are all used to lay out advanced process production lines below 18 nanometers."

"The No. 500 plant, which started construction two years ago, has begun to undertake various tests of the -nanometer process. We are currently solving many problems and improving the yield rate. We strive to complete the formal mass production in June and supply the group with -nanometer process for the production of core product S chips."

"And according to our reliable intelligence, TSMC's 20nm process is expected to enter mass production in the second quarter."

"Factory 18, which started construction last year, is mainly used to undertake the new generation of 14-nanometer technology. The equipment installation has been completed and the final stage and equipment debugging are in progress. It is expected that after April, it will be able to cooperate with the process R&D department to carry out various tests for the 14-nanometer process. It is expected that the 14-nanometer chip will be trial-produced by the end of this year, and mass production is expected to be planned in the second quarter of next year."

"The Plant 21 project, which was launched in the first half of this year, is mainly aimed at the future 10-nanometer process. As a test plant in the research and development stage of the 10-nanometer process node, it is expected to conduct preliminary trial production and testing next year, and mass production in the second quarter of the following year."

"The 17nd plant, which will be launched in the second half of this year, is aimed directly at the future -nanometer process, and will strive to carry out trial production the year after next, and achieve mass production the year after that, that is, in ."

"Factory No. 19 and Factory No. 20 in Jiangcheng Storage. Factory No. 19 mainly produces advanced memory technology, while Factory No. 20 is a flash memory factory."

Today's meeting mainly discussed the field of logic chips (CPU/GPU, etc.), so Ding Chengjun did not say much about the field of storage chips (running memory and flash memory). He just mentioned it briefly and then focused the topic back to the field of logic chips.

Ding Chengjun continued, "Based on the high correlation between process research and development and the factory, we adopt a model of construction and research and development at the same time to ensure that after our future new generation of technology has a breakthrough, the factory will be able to make corresponding production preparations."

"Instead of the current situation where our factory construction lags behind the speed of our process research and development!"

"After all, the current technological advancement speed is relatively fast, and the factory must be built in advance, otherwise it will be too late."

"Our R&D team believes that in the next few years, as technologies such as multiple exposure technology and 3D transistor technology gradually mature, the development of the process will be rapidly promoted."

"Our R&D team, especially Professor Liang Song, believes that by the end of this year at the latest, we will be able to perfect the 14-nanometer process and then start various tests to prepare for mass production. We expect to be able to put it into production in the first half of next year. The overall technological progress is relatively fast."

"The year after next is expected to be further advanced to 10-nanometer technology."

"The year after next, we will push the process to 7 nanometers!"

"Because the above series of process advances are based on the deep exploration and advancement of existing multiple exposure technology and 3D transistor technology, and are limited by the physical limits of lithography machines, the overall advancement of transistor gate length technology is relatively small, perhaps just a tiny improvement of two nanometers or even one nanometer."

"The more important process progress is still based on the design of 3D transistors and the improvement of multiple exposure technology, and our process R&D department has great confidence in this regard."

"Our technical team believes that we will not encounter too many critical technical problems before advancing to the equivalent of 7 nanometers. On the contrary, the process advancement after the equivalent of 7 nanometers will be very troublesome, but this is not a problem of the process, but a problem of the lithography machine."

"In the early stage of our 7nm equivalent process development, we are going to use the existing DUV immersion lithography machine for multiple exposures. However, this method has low production efficiency, low yield rate and high cost."

"The technical department believes that if we want to achieve a process lower than the equivalent of seven nanometers, it is better to use EUV lithography machines with shorter light source wavelengths. If we continue to use DUV immersion lithography machines, it will be difficult to accept the cost and commercialization will be difficult."

"Of course, there is no EUV lithography machine now, and it is too early. We will not talk about future processes smaller than 7 nanometers for the time being. We are currently focusing on 18 nanometers, 14 nanometers, 10 nanometers and the future 7 nanometers, which can be produced based on the existing DUV immersion lithography machine."
"Based on the rapid progress of process technology, we also need to make advance preparations for corresponding advanced production lines, so investment and construction of new projects is necessary."

"After all, the improvement of technology is inseparable from the improvement of chip factory technology, and they complement each other. Otherwise, it will be useless if there is technology but no factory."

Others including Xu Shenxue also agreed with this statement. They had suffered the loss this year and must not continue to suffer such a loss next year.

The progress of Zhiyun Microelectronics' 18-nanometer process is lagging behind. In addition to the difficulty of the process technology itself, it is also related to the fact that the project progress at the factory cannot keep up.

At the headquarters of Zhiyun Microelectronics, the No. 13 factory in the Shenzhen base, which was planned to be used for the -nanometer process, completed the equipment installation work in October , and the subsequent finishing and debugging took nearly three months.

It was not until January this year that we started to cooperate with the 18-nanometer process project team to carry out various tests and adjustments to the 18-nanometer process. In February, that is, this month, we started formal trial production, but the yield rate is not very good. Large-scale mass production will have to wait until after June.

The advancement of process nodes, from the research and development of process node technology, to adjustment and testing, trial production, improving yield, and then large-scale mass production, each stage takes time.

If the first few steps take too long, it will be difficult to catch up later.

Ding Chengjun said that Shenzhen Plant No. 17 will be able to start production of the 18-nanometer process in June. This is the result of a large number of technicians working overtime every day and risking their lives...

So at this time, Xu Shenxue said: "We would rather let the factory wait for the process than let the process wait for the factory, otherwise it will have a huge impact on the group's overall strategy, so the plans for Factory 21 and Factory 22 must be strictly implemented according to the plan and cannot be delayed."

When Ding Chengjun heard this, he felt a little relieved. He knew very well that the investment in each factory here was a huge investment starting from several billion US dollars. Without Xu Shenxue's support and nod, it would be impossible to go ahead.

Now that Xu Shenxue has personally expressed his support, it will be much easier to obtain funding and investment in the future.

Ding Chengjun, feeling a little more relaxed, also decided to bring some good news to other senior executives of the company, so that they would not always feel that Zhiyun Microelectronics was spending more money than it could earn back, spending so much money every year.

They, Zhiyun Microelectronics, can also make money!

Ding Chengjun said: "After we started mass production of the 28-nanometer process last year, we have continued to explore the technology and improve and upgrade the 28-nanometer process, and we have achieved initial results."

"We have launched a new generation of 28nm low-power technology, namely LP process. Compared with the first generation nm process technology used before, it can reduce power consumption by % under the same performance, which has quite good market competitiveness."

"This process has been first applied in Tongcheng Plant 28 and Shenzhen Plant , with a monthly production capacity of pieces. At the same time, we are also upgrading the technology in Tongcheng Plant , Shanghai Plant , and Shenzhen Plant , so that these three plants will also have the production capacity of this process, increasing the production capacity of the LPN process to a monthly production scale of pieces."

"The 28LPN process is currently used in the production of various new-generation 700-nanometer chips of the group, including the new-generation W 3001 series chips, AI1100 chips, OPA1500, NVIDIA GPUs, as well as SOP2 series server chips and SOP P series CPU chips."

"The improved 28-nanometer process chips launched by the group this year will all adopt our new generation LPU process, which can effectively enhance the market competitiveness!"

"At the same time, we have also taken on some external orders, including Huawei's SOC and chips from some other domestic chip manufacturers. At the same time, some chip manufacturers are in talks with us, and we expect to get more orders this year."

As he spoke, Ding Chengjun took out a revenue estimate chart:
"Due to the strong market competitiveness of the 28LPN process, the -nanometer process has become the core revenue source of Zhiyun Microelectronics, accounting for percent of the company's revenue."

"In the global market, our 28-nanometer process accounted for about 30 percent of the market share of this process node in the first quarter of this year."

Xu Shenxue naturally knew why... The internal orders of the group alone were enough to support this process node, with more than 30% of the global market share.

Why is Zhiyun Group determined to engage in semiconductor manufacturing? It is because its own demand is too huge. The shipment volume of various chips using Zhiyun Group's 28-nanometer process alone accounts for about 30% of the global market share.

This resulted in a large number of chip foundry orders from the group, which directly pushed Zhiyun Microelectronics to the throne of the world's third largest chip factory. It also gave Zhiyun Microelectronics sufficient confidence to continue to invest heavily in the research and development of advanced processes.

Ding Chengjun continued, "Because the technology is quite mature, our OEM gross profit margin is relatively high, and we have sufficient orders. This year's overall capacity utilization rate is expected to reach more than 90%, and the overall revenue and profit will be very good. Compared with last year, the revenue in the first quarter of this year increased by 36%, and the profit increased by 23% year-on-year."

"Based on the current market changes and those in the next few years, we predict that the 28-nanometer process, especially the low-power 28-nanometer process, will continue to see strong demand. In addition to our internal orders, there will also be a large number of external orders."

"The market demand for 28-nanometer technology is extremely strong, and the market size will be huge in the coming years. We will continue to promote the technological transformation of some existing old factories, and upgrade the upper limit of these factories from nanometers, the old generation of nanometers to the new generation of LPN technology."

"For the 45-nanometer process, we have utilized many existing mature technologies to improve the low-power 40-nanometer process technology. Compared with the 45-nanometer process technology, this process has lower power consumption and better performance, but the production cost has not increased. The overall cost performance is relatively high. At present, this technology has been gradually applied and accepted orders for mass production."

"The 28nm and 40nm processes will require little investment and yield big returns in the next few years. At the same time, the market competitiveness of low-power technology will also be greater, which is expected to further boost our company's revenue and profits."

At this time, Ding Chengjun smiled and said, "We have worked hard for several years, and now it is time for us to reap the fruits of our labor."

Ji Chenghe said, "It seems that your advanced process is quite good. The low power consumption is quite good."

Xu Shenxue also said: "For chip manufacturing, more advanced technology is the core competitiveness, even at mature process nodes. Why do customers choose us instead of other companies for the same 28-nanometer process? We always have to give customers a reason to convince them, whether it is more advanced process technology or lower prices."

"Compared to lower prices, the use of more advanced technology is undoubtedly more advantageous and can also bring more lucrative returns."

The chip foundry industry is inherently capital-intensive, with huge investment scales and very high equipment depreciation costs. At the same time, there are many manufacturers with mature processes, resulting in fierce competition.

At this time, if there is no unique technological advantage and you just simply engage in price wars, it is actually difficult to form a virtuous circle.

Without profits generated by a virtuous cycle, what will drive further development of the technology?
At this time, Ding Chengjun continued: "In addition, there is still a market for nodes above 65 nanometers. Our mature process department also continues to conduct in-depth research and development on processes above 65 nanometers, using a large number of existing mature technologies to reduce power consumption and improve performance, further enhancing the market competitiveness of our process nodes above 65 nanometers."

"We expect that within three years, by expanding the production capacity of several existing 8-inch factories, we will increase the production capacity of 65-nanometer and 90-nanometer processes by another 50%, and the production capacity above 130-nanometer by another 30%, so as to meet the growing demand for mature process technologies in China."

"This is also to support domestic semiconductor core equipment and consumables manufacturers such as Gulf Technology. After all, we are the largest, and even the only, customer of some domestic equipment and consumables. Therefore, we will not give up on nodes above 65 nanometers, but will continue to increase production capacity."

"This part of the support plan is expected to be completed in three years. After three years, our company will no longer expand production capacity above 65 nanometers!"

"Because there are already many chip manufacturers in China investing in this field, according to our observation, domestic chip manufacturers represented by Central International are investing heavily in building new factories. In order to save costs, a considerable part of their new factories are purchased from Fairy Mountain Holdings' various equipment and consumables."

"The orders from these domestic manufacturers and even some foreign small chip manufacturers are enough to maintain a virtuous cycle for Fairy Mountain Holdings."

"Three years later, our support targets should be the new generation of semiconductor core equipment and consumables represented by DUV immersion lithography machines."

Ding Chengjun talked a lot about Zhiyun Microelectronics' investment plan and project R&D plan, then turned to the last page of the PPT and began to reveal his true intentions.

He pointed to the huge number on the top and said, "The expansion of the 28-nanometer process and 65-nanometer process capacity is expected to cost about US$3 billion within three years, mainly for upgrading existing factories. At the same time, a new 12-inch wafer fab for mature use and two 8-inch wafer fabs will be built."

Three billion US dollars is not taken seriously by the executives of Zhiyun Group, not to mention that the three billion US dollars is divided into three years of investment, which is only ten billion US dollars a year. It's no big deal!
Moreover, these mature technologies will bring a lot of revenue and profits to Zhiyun Microelectronics in the next few years. This is a business that can be said to be very profitable at first glance.

But what Ding Chengjun said next left everyone present speechless.

"We expect to continue to invest approximately $30 billion in process R&D and factory construction for advanced process nodes above 18 nanometers in the next three years, most of which will be used for the construction of new factories, and a small part will be used for process technology R&D."

Hearing the huge figure of 30 billion US dollars, many people present frowned.

This is no small amount. Although it is data from three years, it averages out to 10 billion US dollars a year.

Kong Weiquan, the group's senior vice president and chief financial officer, said, "Isn't this investment too big? Thirty billion US dollars in three years is not a small amount. Once invested, it will probably further reduce the overall profit level of our group."

"At the same time, what is the expected revenue and profit of the advanced technology? Can the expected goals be achieved? Not to mention making money, can the depreciation of this equipment be spread out?"

"As far as I know, the depreciation rate of advanced semiconductor equipment is very high. TSMC's depreciation rate is over 35%, and your microelectronics' depreciation rate is 32%."

"With such a huge initial investment and huge equipment depreciation rate, can the profits brought by these advanced technologies be sustained in the future?"

At this time, Ding Chengjun said: "According to our forecast, our expected revenue and profit are fully able to support its initial investment and equipment depreciation rate."

"At present, our chip foundry revenue comes from internal orders within the group, and 80% of the internal orders are orders for advanced process technology. In particular, the demand for advanced processes in the field of mobile phone SOC is endless. At the same time, the CPU and GPU fields are also continuously pursuing advanced processes."

"According to my estimation, even if the group's semiconductor business does not grow at a large scale, the current market size is sufficient to support most of our advanced process production capacity."

"According to our calculations, our integration costs for investing and producing chips ourselves are more advantageous than looking for TSMC to do the manufacturing."

At this time, Xu Shenxue also said: "It is normal for semiconductor investment to be large, but we cannot shrink our hands and feet just because the investment is large. We still need to develop this advanced technology. Not only must we develop it, but we must also increase our efforts to develop it." "30 billion US dollars in three years, invest!"

Xu Shenxue has always been a staunch supporter of investment in the semiconductor field. Not to mention $30 billion, even if it was $40 billion or even $50 billion, he would dare to grit his teeth and invest in it.

When it comes to spending money, Xu Shenxue is not afraid of anyone!

In a semiconductor strategy meeting, Zhiyun Group, with the personal approval of Xu Shenxue, finalized a massive investment plan totaling US$33 billion over the next three years to build several chip factories and promote the advancement of process technology.

Not only in semiconductor manufacturing, but also in this meeting, follow-up investment plans in the field of semiconductor design were finalized.

Zhiyun Group's semiconductor design business is centered around Zhiyun Semiconductor Co., Ltd., which has multiple business units, including the SOC Business Unit, SOP Business Unit, GPU Business Unit, Storage Business Unit, and several important subsidiaries, such as NVIDIA, Vision Technology, and Pangu Technology, as well as dozens of subsidiaries engaged in various auxiliary and functional chip design businesses.

At the meeting, in order to continue to maintain Lingyun's leading position in chip design, high investment will continue in the next few years to ensure the leading position of several of the company's core products, mainly the S series chips and W series chips used in smart terminals. At the same time, key technical research and development will be carried out in the GPU and CPU fields.

For example, to ensure the smooth development and market launch of X86 CPU, all PC products under the X86 architecture in the group are now waiting for the group's self-developed X86 chips.

Therefore, it is still necessary to ensure investment in the field of semiconductor design. However, some auxiliary and functional chip businesses were also cut at the meeting. These were done before because the domestic market did not have them or could not meet the requirements, so we had to force ourselves to do them.

However, in recent years, with the rapid development of the consumer electronics industry chain, many chip design companies have emerged in China, and many of them are doing quite well. In this case, Zhiyun Semiconductor has cut off some non-core products that have domestic alternatives, and reduced R&D investment in product lines and non-critical products by selling technology and business subsidiaries, or merging with some domestic chip companies, so as to concentrate its efforts on core projects.

Both semiconductor design and manufacturing are very expensive. After the meeting, when Ji Chenghe and Xu Shenxue were drinking tea and chatting in the office, Ji Chenghe sighed: "The semiconductor manufacturing field is too expensive. We work hard to sell mobile phones to make some money, but in the blink of an eye, it's all spent by these semiconductor people!"

Xu Shenxue took a sip of tea, moistened his throat and said, "This can't be helped. Semiconductors are the core lifeblood of our group, so we have to bite the bullet and do it no matter how much we invest."

The semiconductor field is directly related to the core performance of Zhiyun's various intelligent terminal products. It also affects PC products as well as servers, supercomputers, AI and a host of other things.

The semiconductor manufacturing field is the core foundation of a large number of terminal businesses for Zhiyun Group... It may not make money or even make losses, but it cannot exist.

Not only must we have it, but it must be extremely advanced, striving to be on par with our main rivals, or even ahead of them.

Otherwise, Zhiyun's mobile phone chips use 28 nanometers, while competitors such as Apple use 20 nanometers or even 14 nanometers chips. How can they compete?

But fortunately, not long after this meeting, Ding Chengjun of Zhiyun Microelectronics brought good news to Xu Shenxue.

"Mr. Xu, our 18-nanometer process production line has just conducted a small-scale trial production at Shenzhen Plant 17. Except for the yield rate which is still a little insufficient, all other performance data indicators have reached the design requirements and have reached world-class standards.

Ding Chengjun seemed a little excited when he gave a report to Xu Shenxue: "This world-class level is not just our boasting, but a real world-class level."

"One of the most core technical indicators is transistor density. Our 22-nanometer process has reached 3 million per square millimeter. This data is roughly the same as Intel's -nanometer D transistor process, and slightly exceeds TSMC's -nanometer process of million per square millimeter, if our intelligence is correct!"

"In addition to transistor density, we also have a low power consumption advantage over TSMC's 20-nanometer process. Lower power consumption means we can make the chip's main frequency higher, improve the overall chip performance, and control heat at the same time."

"When we used the S503 chip for trial production, the S503 chip produced by the test was very outstanding. Compared with the S403 chip produced last year using the -nanometer process, the CPU performance has been improved by at least %, and the GPU performance has been improved by %."

"Not only that, we have also reduced the size of the chip and reduced power consumption, which has greatly improved the overall performance of the chip."

Ding Chengjun said: "After trial production, it has been proven that our 18-nanometer process has fully met the design requirements. Strictly speaking, it has even exceeded some of our predetermined technical indicators."

"The only drawback is that the cost is slightly higher than last year's S403, but we are working hard to improve the yield rate and strive to raise the yield rate to an acceptable level by June and control the cost to the level of last year's S403."

As Ding Chengjun spoke, he also handed over a technical data report.

Xu Shenxue was reading the report while listening. Just by comparing the various technical parameters in the technical report, it is easy to see that Zhiyun's 3-nanometer process is actually similar to Intel's -nanometer D transistor process. The performance data in all aspects are also relatively consistent, including the gate length and transistor density, which are actually the same.

After all, the two belong to the same process node, so it is reasonable that their performance is similar.

As for why Zhiyun Group insisted on using the term 18-nanometer process, it is purely for commercial promotion purposes... Who asked TSMC to insist on coming up with a 20-nanometer process?

Intel, Zhiyun and TSMC, these three chip manufacturing companies' process technologies are currently in the first tier.

If we compare them from the perspective of comprehensive performance, Intel is still the leader. Their 3nm D transistor process has become more powerful than when it was first launched after mature iterations.

The second is the 22nm process (3nm D transistor process) that Zhiyun has just trial-produced. It is slightly weaker than Intel, but the gap is not that big.

Next is TSMC's 2nm process. In fact, their technology is quite advanced. They have achieved a transistor density that is close to that of Zhiyun/Intel's 3D transistor process with D transistors. Although it is slightly less, the difference is not obvious and can basically be considered to be at the same level.

However, compared with 2D transistors, 3D transistor design consumes more power while maintaining the same performance.

This means that if TSMC's 20-nanometer process and Zhiyun's 18-nanometer process are used to produce the same chip, the chip produced by TSMC's 20-nanometer process will consume more power and generate more heat.

This problem is actually not a big problem in the PC field, because there are fans for heat dissipation, and at worst water cooling is enough...

However, in mobile phones, where every inch of space is valuable, this brings considerable trouble, as there is not enough space in the phone to design and install an additional cooling system.

According to intelligence obtained by Zhiyun Group's intelligence department, in order to adapt to TSMC's 8-nanometer process technology, Apple even limited the chip's main frequency when designing the A chip. This is to reduce the main frequency to avoid excessive power consumption and heat problems... In other words, it was not an intentional restriction, but a targeted design based on the process level.

In this way, when Zhiyun releases a new generation of flagship mobile phones in the second half of this year, it will once again take the lead in chip performance.

The transistor density of the two seems to be similar, but the A8 chip can only play at a low-power version in order to limit heat, but the Zhiyun S503 can increase the main frequency and performance without any scruples. Here comes the gap!

Just by looking at the comparison of these data parameters, Xu Shenxue was able to foresee a scenario in which his own mobile phone would outperform the other party in terms of performance after the flagship phones of both sides were launched in the fall.

Of course, no matter what, the Fruit A8 chip, which uses TSMC's nm process, will still have very strong performance. It is the core competitor of Zhiyun chip this year, and it can even be said to be the only competitor.

As for other competitors, namely Qualcomm, MediaTek, and Samsung's SOC chips, this year Zhiyun's S503 can beat them all!
The reason is very simple, they can only continue to use the 28 nanometer process!
Although Qualcomm has a new generation chip plan with 20-nanometer chip process, it is impossible for them to snatch TSMC's 20-nanometer process production capacity from Apple.

This means that if Qualcomm wants to ship its new generation Snapdragon 20nm process chips on a large scale, it will probably have to wait until next year. How can it compete when it doesn’t even have the goods?

MediaTek is in even worse shape. Although it has entered the mobile phone SOC field, its survival is difficult because the domestic low-end market has long been taken over by the Zhiyun W series. It is also impossible for them to grab TSMC's 20-nanometer advanced production capacity.

Four Star's self-developed SOC chips also face the embarrassment of not having access to advanced technology. Most of this year's chips can only continue to use Four Star's 28-nanometer process, and it is estimated that they will not be able to get 20-nanometer chips until early next year.

This year's mobile phone chip competition is still between Zhiyun's S500 series and Fruit's A8. The others can only watch from the sidelines!
This chip competition, to some extent, is also a battle between Zhiyun Microelectronics' 18-nanometer process and TSMC's 20-nanometer process.

As for several other chip design companies, such as Qualcomm, Quadcore, and MediaTek; chip manufacturers, such as Quadcore, GlobalFoundries, and UMC, they are already behind the times.

Among them, Quadcore's 3-nanometer process progress this year has been delayed, and it is estimated that trial production will not be possible until the end of the year, and mass production will not begin until early next year. At the same time, their -nanometer D transistor process is also very troublesome and has not been resolved yet.

Part of the reason for all this is that they were at a disadvantage in the battle for chip talent in the past two years.
Professor Liang Song, who was previously from TSMC and focused on 3D transistor technology, wanted to change his employer. Sixing contacted him first, but was later intercepted by Zhiyun...

What effect did this lead to?
According to netizens, Four Star failed to recruit Liang Song, so the 3nm D transistor process continued to be difficult to produce.

In contrast, after Zhiyun recruited Liang Song, its technology progressed smoothly. It took only a few years to advance to the 3-nanometer process and adopt D transistor technology.

This is the serious impact that top talent brings...a top talent can really change the entire industry.

Of course, Xu Shenxue knew that the reason why Zhiyun Microelectronics' process advancement was so smooth was not only due to the joining of top talents like Liang Song, but also inseparable from the huge role of the scientific research system.

The two complement each other and ultimately lead to such huge results.

Since they failed to recruit Liang Song and couldn't find other heavyweight chip talents to replace him, Sixing is now having trouble advancing its technology.

It’s not that impossible to complete it, but it will at least be delayed for two to three years.

There is also a major chip foundry, GlobalFoundries, which was previously a chip factory owned by AMD. Later, AMD sold its chip manufacturing business and focused on chip design after getting rid of the burden.

So this part was sold, and the AMD chip manufacturing business was taken over by West Asia Capital and renamed GlobalFoundries. After acquiring other chip factories, it became the second largest chip foundry in the United States.

Their technology is even more backward, not even as good as four-star technology. If they rely on their own research and development, they probably won’t be able to develop it in several years.

UMC is similar. Although it has not stopped its research and development, it can no longer catch up with the pace of the top giants...

GlobalFoundries and UMC have fallen behind, which is, to some extent, their own choice... It is too expensive to continue to follow up on advanced technology, requiring investments of tens of billions or even hundreds of billions of dollars.

The shareholders were all terrified when they saw this. They decided that it would be better to make some money first with mature technology, and then wait until time passed and low-process technology became easier to develop.

If we invest huge amounts of money now to compete with Intel, Zhiyun, TSMC, and Sega in technology, and try to catch up with each other, we will end up dying a miserable death.

Apart from anything else, it is difficult to get orders... In fact, there are not as many large orders for top-level technology as imagined, and they are concentrated in the hands of a few leading chip design companies: Intel, AMD, Qualcomm, Fruit, Zhiyun, Sixin, Warwick, MediaTek and a few others.

There was originally an NVIDIA one, but now NVIDIA has become a subsidiary of Zhiyun Semiconductor, and therefore also belongs to Haizhiyun.

Among these large orders, quite a lot of them are produced and sold by themselves, without looking for OEM at all.

For example, Intel produces all its chips by itself and does not need any foundry.

It’s similar for Zhiyun. The various chips under Zhiyun Semiconductor, especially the two SOC chips, the S series and the W series, have an annual shipment volume of more than 600 million pieces, which is quite exaggerated.

Then Zhiyun itself has a large amount of memory and flash memory chips, as well as CPUs and GPUs in the PC field, including the newly acquired GPU chips from Nvidia.

All the above-mentioned chips under Zhiyun Semiconductor are produced by Zhiyun Microelectronics, and there is no need to find a foundry.

What’s even more interesting is that the chip shipments of the above two companies, which design and produce their own chips, also occupy the mainstream of the market.

Intel is the undisputed leader in the PC and server markets.

Zhiyun Semiconductor is the undisputed leader in the communications market.

This means that most of the global production demand for advanced process chips is met by these two companies self-sufficiently, and there is no need to seek external outsourcing.

No, there is another company called Four Star, they also have chip design business, but they also produce their own chips without outsourcing.

After excluding these three companies, the companies that really need to find top-level process outsourcing are Apple, Qualcomm, AMD, Warwick, and MediaTek, chip design companies that do not have chip factories.

Then TSMC, UMC, GlobalFoundries, as well as Fourstar Electronics and Zhiyun Microelectronics, which, in addition to satisfying their own production and sales, are also looking for large-scale external foundry orders, are all competing for orders from these chip design companies.

Zhiyun has won orders from Huawei, and TSMC has won orders from Apple, Qualcomm and AMD.

As for UMC and GlobalFoundries, there is not much left.

There are not many orders, so there is no revenue or profit, and it is even more impossible to raise a large amount of funds to promote technological progress.

Anyway, the latter companies such as GlobalFoundries and UMC are now lying flat, and they are just thinking about making money with mature technologies. They can just talk about advanced technologies, but if they really want to do it, we can wait a few years.

Even if these second-tier companies want to compete, it’s useless because they can’t win!

In mid-February, Zhiyun Microelectronics officially announced that its 18-nanometer process had passed technical verification and officially launched the mass production plan, which is expected to start in June.

In addition to accepting orders from within the group, they also strive for external orders.
AMD has already expressed some interest in launching their new generation flagship CPU and GPU based on Zhiyun Microelectronics' 18nm process technology.

However, whether the cooperation will be reached depends on the subsequent negotiations. This kind of cooperation between competitors involves many things and cannot be finalized in just a few words.

AMD and Zhiyun Semiconductor, a subsidiary of Zhiyun, are still major competitors in the CPU and GPU businesses on the PC side.

But even so, Zhiyun Microelectronics will not refuse their orders... Let alone AMD, even if Qualcomm or even Apple came to them and asked Zhiyun Microelectronics to manufacture chips for them, Xu Shenxue would agree with a smile and even give them a discount!
Because if Zhiyun doesn't take the order, TSMC or Sixin will...and after TSMC and Sixin have these orders, they will have more funds to promote the development of advanced processes.

If you want to curb the technological development of TSMC and Four Star, the best way is not to play the game of spending money to promote technology with them, but to directly snatch their customers... to cut off their financial support.

A large number of advanced foundry orders have been lost, revenue has been lost, profits have been lost, what else can they use to promote technological development?
If these two companies don't develop their technology, in three to five years, maybe it won't be foreign countries that sanction Zhiyun, but Zhiyun Microelectronics that will sanction foreign countries...

Compared with the competition in the terminal field, Xu Shenxue pays more attention to the leading edge of underlying semiconductor manufacturing.

Leading in semiconductor manufacturing is the real leading!
Unfortunately, Apple and Qualcomm are not following Rational Cloud, at least not for now.

But who can say for sure what will happen in the future?

What if?
Am I right!

(End of this chapter)