Editor’s note: Since the U.S. cut off the supply of ZTE in 2018, the status of the chip industry has been raised to an unprecedented height. China has conducted a “chip breakthrough” from top to bottom. The industry breakthrough is simple to say, it is nothing more than mutual card positions. Let the other party have no neck to get stuck; the industry breakthrough is also very complicated. In addition to the technical shortcomings that need to be broken through, there are too many problems to be solved in planning, market, mode, direction, and capital.
“Intelligent Relativity” attempts to analyze all aspects of “chip breakthrough” from a three-dimensional perspective. This is the third article on the topic of “chip breakthrough”, and the strategic differences of domestic mobile phone brands in the chip field.
The core points of this article:
●The differentiated direction of “Huami OV” chip deployment and the multiple reasons for their persistence and the chip industry
●Can domestic mobile phone brands usher in a “generational leap” in the chip field?
On August 17, US local time, US Secretary of State Pompeo stated that the US Department of Commerce has further tightened restrictions on Huawei’s access to US technology, and at the same time included Huawei’s 38 subsidiaries in 21 countries in the “entity list”.
Industry insiders said that this basically blocked Huawei’s subsequent purchase of chips through other chip manufacturers using US chip technology.
In the book “Core Matters” published in July 2018, Professor Wei Shaojun, director of the Institute of Microelectronics of Tsinghua University, said, “China’s chip technology and industrial development will take at least 20 years and a generation of hard work before it is possible to climb into the world. Peak.”
At that time, the “Meng Wanzhou Incident” had not occurred, the entity list had not yet been released, and the United States had not implemented a complete blockade of Chinese technology companies. The words are too long.
1. The second “twenty years” of Huawei’s core manufacturing
“Wait to whoever misses the crown and cover, it has been hard for twenty years.”
Since 2000, it has been a turning point for China’s semiconductor industry. So far, it can be said that China’s semiconductor industry has officially stepped on the right track. After 20 years, domestic mobile phone brands have sprung up, and domestic mobile phone brands have also become the rise of Chinese chips. Important participants.
After the “one-sheet checklist”, Huawei was forced to a “dead road”. At the same time, the “spare tire plan” was known to the world, and it turned out that Huawei had always kept a hand.
During the accelerated period of the domestic semiconductor industry, Huawei set its sights on the chip field. In 2006, Huawei officially started the research and development of smart phone chips. Of course, this time is by no means too early, so that the first mobile phone chip K3V1 released in 2008 was due to The manufacturing process is too far apart, “being born is death”.
But in fact, as early as 1991, Huawei established an ASIC design center. For Huawei, he has already gone through the first “20 years”. However, what he exchanged was just a basket of cold water. No matter how unwilling, “Xinlu For the first time, the difficulty of “journalism” has also been deeply recognized.
In 2012, Huawei’s second mobile phone processor K3V2 was unveiled. This was known as the “world’s smallest quad-core chip” that year, and it indeed achieved tens of millions of commercial use. But looking back at this time, K3V2 may not be so awesome.
Also in the same year, Huawei confirmed the direction of self-research of Kirin chips. After another “ten years” of time flies, Huawei had to announce that “Huawei’s high-end Kirin chips will be discontinued on September 15.”
The comprehensive blockade in the chip field has caused varying degrees of blows to many of Huawei’s business lines. For example, some of Huawei’s server chips, AI chips, and 5G base station chips use 7nm process technology, so they will face the “less and less used” technology in the future. Embarrassing situation, but the most urgent thing is the mobile phone Soc chip.
It is reported that TSMC is currently full of firepower to help Huawei in chip manufacturing, and in other areas such as 5G base station construction. Because of the upper limit of the number of construction, with TSMC’s monthly production capacity, it can fully accumulate a large amount of “inventory” before September 15. , And Huawei’s mobile phones ship as many as 200 million units each year, which means that more than 200 million kinds of chip products are needed each year.
Huawei made two-handed preparations. At the end of July, Huawei and Qualcomm signed a US$1.8 billion patent settlement agreement, which means that Qualcomm will have the opportunity to return to the ranks of Huawei’s main 5G suppliers.
MediaTek has also been further focused. According to foreign media reports, Huawei mobile phones will use MediaTek chip products on a large scale in the future. At present, it has signed a chip purchase order with MediaTek. With the release of high-end 5G chips such as Dimensity 1000, it is for both parties. Good news, but can this really solve the “worries” of Huawei chips?
Of course, after years of trial and accumulation, Huawei has never achieved nothing in the chip field. The server chip Kunpeng 920, a completely self-produced chip, landed in May last year, breaking Intel’s dominance in the server core field in one fell swoop. In China Mobile’s server procurement project in 2020, Huawei won the bid for more than 40,000 servers, and the winning bid accounted for 30%. Nearly 20,000 of them are equipped with Huawei’s Kunpeng processors.
The Kunpeng series being accepted by the market is definitely a positive sign for Huawei.
Why Huawei is “enthusiastic” in chip development? As we all know, it is only one of the reasons why it has the “right to speak”, “bargaining power” and “have to do”; secondly, Huawei’s “preference” for technical improvements in product details allows for independence. Chips have become a necessity. For example, the “five functions” brought by the Mate 30 last year, such as high-speed photography, low-light shooting, wireless fast charging, side touch, and air operation, are based on the Kirin 990 series chips that Huawei’s Smart phone software and hardware have achieved better coordination, and independent chips can best meet their own needs. This is irreplaceable by any other chip, otherwise it can only be what others sell you and what you use.
The status quo is a challenge as well as an opportunity for Huawei, and Huawei must make a breakthrough in the “next decade”.
2. Can Xiaomi’s “core road” cast a future?
Not long ago, on the tenth anniversary of Xiaomi, Lei Jun said, “Although the surging chip has encountered huge difficulties, it has not given up!” Compared with Huawei, Xiaomi’s chip development is much later, and it is of course natural. After all, Xiaomi, which was only established in 2010, wants to be fast. Can’t get up quickly, but if it is for Xiaomi itself, Xiaomi is not slow to open the road to chips.
In 2014, 4 years after the establishment of Xiaomi, the self-research of mobile phone chips was officially launched. In 2017, the Surging S1 was unveiled, but the effect was not satisfactory. In addition to the technical bottleneck, the S2 has not been able to achieve mass production, and gradually faded out of people. Vision.
Huawei’s history has allowed others to see the hardships of self-developed chips. Measuring time in units of “ten years” and investing funds in units of “billions” is not something that “ordinary people” can persist; the reason why Qualcomm is ahead , Because of the advance layout; Apple’s self-developed chips for many years still have little effect, and now they are still subject to Qualcomm.
Therefore, Xiaomi has gradually changed its strategy. Compared with its own end, will investment be a better way?
Investing in Jingchen Semiconductor and Espressif Systems is a signal. In the past two years, Xiaomi’s investment scale has been further expanded. In April last year, Xiaomi announced the reorganization of its Songguo Electronics team, and carried out a team split to form a new company “Nanjing Dayu Semiconductor”. It has begun independent financing, and its main business is manual labor. Smart chip development is based on the Internet of Things; subsequently, Xiaomi invested in Ankai Microelectronics, a chip company for the Internet of Things.
Starting this year, Xiaomi has increased its investment speed to the extreme. Starting from January 16, Xiaomi has quickly invested in 8 semiconductor companies in just over two months through its “Xiaomi Yangtze River Industry Fund”, which specializes in semiconductors. Companies: Diao Microelectronics, Sutong Semiconductor, Chipset Microelectronics, Fortior (Fortior Technology), Angray Microelectronics, Aojie Technology, Smart Microelectronics and Hanxin Microelectronics.
It hasn’t stopped yet. In August, it first led the investment in Xinlai Technology. On August 10, Tianyan Check showed that Brite Semiconductor (Shanghai) Co., Ltd. had undergone many business changes and investors added Hubei Xiaomi Yangtze River Industry Fund Partnership (Limited). (Partnership), August 17, also from the data from Tianyan Check, Ningbo Gekong Intelligent Technology Co., Ltd. has undergone industrial and commercial changes, and the company’s investors have added Hubei Xiaomi Yangtze River Industry Fund Partnership (Limited Partnership), ranking the eighth largest shareholder.
After these two years of “crazy eye-opening”, Xiaomi has invested more than 20 in the semiconductor field, covering multiple chip product areas from Wi-Fi chips, radio frequency (RF) chips, MCU sensors to FPGAs.
It can be clearly seen that Xiaomi’s advancement in the chip field is obviously not suspended as the outside world believes, and unlike Huawei’s hands-on efforts, “curve core” is undoubtedly the best choice for Xiaomi when its technical capabilities are insufficient.
The underlying reasons for Xiaomi’s “no abandonment, no abandonment” of “chip” products are the same as Huawei, but they are also slightly different. It has not suffered from Huawei’s embarrassment and has become Xiaomi’s biggest embarrassment at present. “No technology and no core capabilities” has always been the best way for the market to mock Xiaomi. Xiaomi is unwilling to carry forward with “infamy”. In Xiaomi’s own words, “chips” It is the commanding height of mobile phone technology, and Xiaomi wants to become a great company, and the primary factor to become a great company is to master the core technology.”
At the same time, it chose to enter the chip field by investment. “Intelligent Relativity” believes that this behavior is also very “Xiaomi”. Throughout the Xiaomi ecological chain, it uses investment to find the best team and use Xiaomi’s platform and resources. By rapidly deploying related industries, Xiaomi has formed a set of mature and effective methodology.
When Xiaomi enters the third phase of its strategy, that is, when it breaks through the existing development bottleneck with the Internet of Things strategy, the chip has indeed become a hurdle that Xiaomi cannot bypass. The “Xiaomi Eco Model” may also be a way out.
3. The path of “OV” is different
OPPO may be the player closest to Huawei. “Independent core manufacturing” has become OPPO’s strategic direction, and it will go back to 2017.
In 2017, OPPO established Jinsheng Communication, which is a 100% holding company, and the outside world began to speculate whether OPPO will also make cores? It wasn’t until OPPO released many chip design engineer positions on the recruitment website in 2018 that Jinsheng Communications included “integrated circuit design and services” into its business projects, and finally made it clear for OPPO to make cores.
Two months later, OPPO established the chip TMG (Technical Committee), and then OPPO opened the road to digging corners in major companies, first from Spreadtrum, HiSilicon, and MediaTek to dig a group of grassroots engineers; and from MediaTek Zhu Shangzu, the former chief operating officer, and Li Zonglin, the general manager of MediaTek’s wireless communications business department, were successively recruited. So far, OPPO’s mobile phone chip research and development team has taken shape.
In February of this year, a heavy news came from an internal article published by the special assistant to the CEO of OPPO. Before the “Mariana Project” was officially unveiled, Mariana was the deepest trench in the world, with a depth of 6-11 kilometers, and it was also on the earth. One of the areas with the worst environment, just by hearing the name, OPPO knows what a huge “pit” it faces.
The outside world is not very optimistic about OPPO’s core move. Some analysts believe that “if OPPO is not determined to invest at least three to five years, the cost of investment may not be effectively recovered. This is for the company’s operations. Not a small burden.”
There is also a more straightforward voice that “OPPO really does not have a self-developed gene. After all, if you really want to self-study, then the Xiaobawang of the year should be an independent processor. If not, the OPPO MP3 behind should also be self-produced. Processor. No longer, the 2G era has a better chance of ending. The worst, it should end in the 660 era when OV is in the limelight.”
“Intelligent Relativity” believes that although there is no inevitable connection between “brave” and “success”, OPPO should not be underestimated at the initial stage. According to OPPO’s plan, it plans to invest 50 billion yuan in manufacturing in 2020-2022. core. On average, nearly 16.6 billion will be invested in core manufacturing each year, which is equivalent to an increase of 2/3 on the basis of R&D investment in 2019. Obviously, OPPO’s going out is a big step.
Finally, it is VIVO, which is the same “curve core” like Xiaomi, but it is not an investment but a joint venture. VIVO seems to be the slowest on the “core” road, but there is another world.
Last year, working with Samsung to make VIVO’s chip path on track, Samsung’s first dual-mode 5G AI chip, Exynos 980, is the final product of the two companies’ joint research and development time for nearly 10 months.
In May of this year, it was revealed on the Internet that VIVO had applied for two new trademarks, namely vivo SoC and vivo chip. These two trademarks belong to the ninth category, and the product categories covered include central processing units, modems, computer chips, printed circuits, Computer storage devices and so on.
Regarding VIVO’s chip layout, most people’s views are consistent with those of OPPO’s chip layout. They are mainly planning the future ecology, and what it can finally be made will take the test of time.
The logic of the “OV” layout chip may be different from that of “Huami”. If Huawei is to survive, Xiaomi needs to complete self-breakthrough, “OV” may be forced to advance, and foreign chips are already unreliable. Obviously, once Huawei or Xiaomi successfully completes a fully autonomous chip, will it be sold to the competitor “OV”? The answer is no, Samsung is a precedent, so in this context, OV has to move, considering that the future competition does not advance or regress, “OV” does not want to become the next Nokia.
So far, Huami OV has been integrated into this chip war in its own way. If they are given another 20 years, can the chip become their core weapon?
4. Can domestic mobile phone brands complete the “intergenerational leap”?
When it comes to chips, I have to mention Intel. From writing the “truth” of Moore’s Law in the field of chips to the outstanding performance of the Pentium series, Intel, known as the “king of chips”, has witnessed the ups and downs of the entire chip market. It also means that running first may not always lead.
Recently, Intel announced that its 7nm chip plan has shifted from previous expectations by about six months. The current trend is about twelve months below the company’s internal target, and the rising star TSMC 5nm has already started mass production this year.
Wei Qing, the current CTO of Microsoft China, also proposed a concept called “Generation Leap”. He believes that “Almost no company in business history can lead in two consecutive eras. The second era is actually a reference to the first era. Subversion, this requires not just the next step, but the next few steps. Microsoft is a leap from generation to generation, missed the mobile era, but won the cloud era.”
The cases of TSMC and Qualcomm also show that it is not a fantasy to come from behind in the chip field. Moreover, the “critical point” in the manufacturing process has also brought unique opportunities to domestic mobile phone brands.
In terms of the chip manufacturing process, chips are mainly divided into four links: design, manufacture, packaging, and testing. At present, the main focus of the market is on chip manufacturing, but 5nm is already approaching the limit of manufacturing accuracy. It may not take too long to make a breakthrough in the precision of chip manufacturing.
Therefore, it is too limited to focus on the manufacturing process. For example, the chip packaging process has become a new hot topic at this stage. The emergence of “3D packaging technology” may bring breakthrough changes to the industry. Intel, which has become a bottleneck, is also aggressively deploying in the packaging field, and it is very likely to take this opportunity to turn itself over.
There is a very old paragraph about the investment logic of Jews and Chinese. It is said that a Jew opened a gas station and the business was very hot. Other Jews would set up restaurants, hotels, supermarkets, etc. around the gas station. Shopping malls and so on, a set of commercial ecology is established; if it is replaced by a Chinese, then there will be other gas stations built around the hot business gas station.
Although this story is not new, it also reminds us of the importance of “vision”. Returning to the chip field, domestic companies are very lagging behind in all aspects except for Huawei HiSilicon’s advantage in chip design. The reason lies in our country. The overall chip manufacturing industry chain, including upstream raw materials and production equipment manufacturing, is still very weak, and many links are even just starting.
For domestic companies, this is actually an opportunity. It is not necessary to have the ability to design and produce to gain a firm foothold in the chip field. The same is true for mobile phone brands. Chips can also be built from a single field, or production process, or material process, or packaging process, or cutting process and lithography circuit process, even from point to surface. shortcut.
Even direct overtaking is not impossible. Just like the “carbon-based chips” that the Chinese Academy of Sciences has been studying, also known as “graphene chips,” carbon-based chips have many advantages over silicon-based chips. For example, it is highly malleable and can be folded. The chip’s conduction speed is also very fast, and its comprehensive performance is more than ten times that of silicon-based chips. The main research conducted by the Chinese Academy of Sciences is the only one in the world.
Can domestic mobile phone brands complete the “generation leap” in the chip field? “The next ten years” are all looking forward to it.
Some netizens joked that “just like China’s re-entry into the World Cup, the birth of a fully autonomous chip has become another dream of hundreds of millions of people.” However, “Intelligent Relativity” believes that the latter should be much easier than the former, and the big country’s “chip dream” Is accelerating.