After 20 years of development of domestic mobile phones, the pattern has been generally stable, and everyone is no longer keen on marketing and price wars, but has a higher pursuit and pursues the real core things. On the surface, the industry is not so lively, but in fact it has a more difficult task and mission.
Wen | The King of Furong
Editor| Zhao Yanqiu
Not long ago, Apple held a spring 2020 new product conference, if it is not written on the background of 2022, many people still think that this is a few years ago, because mobile phones and iPads are previous models, mold parts do not have to be replaced, only need to change a chip, dare to push to the market, there are many consumers to pay.
You say you have an AMOLD run 870,000, Apple says I have an A15 chip; you say you have a DXO score of 130, Apple says I have an A15 chip; you say I have an AMOLED curved screen, Apple says I have an A15 chip.
This kind of treatment is difficult for other brands to enjoy.
In the mobile phone market with a relatively stable pattern, we have found that the ultimate competition in the industry is still the competition between chips and systems. Looking at the development history of domestic mobile phones in the past two decades, almost all manufacturers did not feel that it was important to make chips before, and Huawei and ZTE also considered from the inherent needs of cost saving. But after a decade or two of historical process, making chips has now become a strategic need for high-level innovation.
Nowadays, almost all mainstream mobile phone manufacturers in China are moving in the direction of chips. The mobile phone industry is also driven by the market, in the step by step self-awakening, the pursuit of the real core of things.
1
On the last day of 1998, the former Ministry of Information Industry and the State Planning Commission issued "Document No. 5", stipulating that mobile phone production should be licensed, setting a threshold, 9 enterprises got licenses, and foreign capital and joint ventures were not eligible to apply. Later, the state not only gave the policy, but also allocated 1.4 billion yuan to support mobile phone manufacturers.
The state is still quite protective of the early mobile phone industry. Foreign capital can not come in, the market demand is booming, and it is too profitable to make mobile phones that year.
With the A8 mobile phone, Xia Xin made 1.1 billion yuan in two years, making a loss-making company the most profitable company in A shares. TCL mobile phone sales revenue in 2002 is 8.2 billion yuan, profit is 1.2 billion, the entire TCL Group relies on mobile phones to raise. At that time, Li Dongsheng was eager to pick up Wan Mingjian, the person in charge of TCL mobile phones, and kissed him twice. Because of the good performance, Wan Mingjian was once a bit fluttery, and those who sold mobile phones all circulated a sentence: "Domestic mobile phones rely on Wan Zong."
In 2002, the hungry Lenovo spent 90 million yuan to buy a stake in Primavera Mobile phone and officially entered the mobile phone market. Although it was early in the morning and a late set, compared with the meager profit of the PC business, the profit of a mobile phone exceeded 1,000 yuan, and drug trafficking was estimated to be almost the same.
In that era, the playing style and routine of the mobile phone industry were the same as those of the previous home appliance industry, which were basically price wars and marketing wars, and the industry competition did not reach the chip level at all.
TCL spent 10 million yuan to sign the Korean beauty Kim Hee Sun at that time, and the advertisement was also a big brand, his name was Zhang Yimou. Nowadays, Jin Xishan has passed, and TCL's mobile phone has also fallen, but he did not expect that the old strategist has soared.
In addition to asking beautiful women to endorse it, the reason why TCL was very popular at that time was that there was a gem set on the flap. The so-called gemstones are actually Southeast Asian jade worth a few dollars, which is similar to the quality you buy at tourist attractions, but it gives people a very high-end illusion.
Wan Mingjian later went to Changhong, also according to the gourd painting scoop, tied a gold-plated belt on the mobile phone to sell for 1980 yuan, and a mobile phone with a pure gold belt sold for 5000 yuan. In order to reflect the high-end temperament, Changhong also specially let Zhiling's sister record a ringtone, "I wish you a long rainbow in your career and a long rainbow in popularity."
Because everyone asked celebrities to endorse them, celebrities were once insufficient. Fu Mingxia endorsed the first letter, Liang Yongqi endorsed Ke Jian, Zheng Xiuwen endorsed Dibit, Li Wei endorsed Bo Dao, Zhao Wei endorsed Xia Xin, Zhang Ziyi endorsed Southern Hi-Tech, Zhou Xun endorsed Yi Mei, and Jet Li endorsed ZTE. Who is not red, look at the mobile phone brand endorsement to know.
It was a barren era, and sales relied on design, various marketing and channels, and did not rely on self-developed chips at all.
What's more, Nokia, Motorola and other international brands are not using their own chips, and no one in domestic mobile phones has to consider the problem of their own cores. But the effects and sting of the chip are already there.
At that time, most of the mobile phone baseband chips came from Texas Instruments, not Shandong Texas, but Texas, USA. Texas Instruments once accounted for 80% of the global GSM baseband chip share, and Qualcomm, as we know it now, was a young brother at that time.
Once the business is monopolized, it is easy not to think of making progress. Texas Instruments' solution requires mobile phone manufacturers to have strong research and development capabilities, and cannot be used directly. The Nokia mobile phone chip is basically from Texas Instruments. However, the cooperation between the two sides has a premise, texas instruments chips must not only be given priority to Nokia, but also a year later before they can be sold to other manufacturers.
The pain of the chip could already be felt very deeply at that time, but due to the disparity in strength, almost no manufacturers wanted to develop their own chips.
One of the things that can be found is that Changhong launched a high-end mobile phone in 2010 and invited Sun Honglei to endorse it. Its full name is Jin Changhong E power A2 Business Class. I don't know, I thought it was a high-speed rail business seat, but people are actually a mobile phone name. But it claims to use The Changhong dual-core quantum core technology, according to its introduction, this quantum core is somewhat similar to the ISP (image processor) that everyone does now, and the main chip is not self-developed.
Later, the so-called quantum core was basically used on the TV screen, and the domestic chip was still not seen in the mobile phone. Changhong, TCL and Konka, these home appliance manufacturers have semiconductor companies, but their chips mainly serve TVs, the threshold is lower, mainly image-related, and has little to do with mobile phones.
Fortunately, MediaTek's intimate little cotton jacket not only provides a complete set of chips, but also provides software, and can be directly customized according to customer needs. MediaTek's turnkey solution has greatly reduced the threshold for making mobile phones. Cottage mobile phones have blossomed everywhere, and MediaTek has also gained the reputation of the king of cottage machines and become the king of Taiwan stocks.
Wan Mingjian's view represented the idea of domestic mobile phones at that time: chips were just a matter of division of labor in production, and as long as there was no monopoly on chip supply, it was not terrible. Don't think that you can't produce chips is that you haven't mastered the core technology, and there is no need to start from the initial stage.
At that time, there was an article analyzing why TCL did not make cores, and the conclusion was very direct: core making is not necessary. The reason seems to be very good: the chip is a high-end technology, and it will not produce results in the short term. And the chip industry is fully competitive, and chip manufacturers will never provide you with products for ideological things such as "national interests".
This kind of thinking has historical limitations, after all, the United States was still very confident, and China had just joined the WTO, opening the door to the arms of the world, and global economic integration was the mainstream.
Suddenly looking back, I found that the world had changed. The post-80s and post-90s textbooks are all about globalization and fine division of labor, and the current hot words are autonomous controllability and localization. Epidemics and wars are profoundly changing humanity, and this change has not only made life unrecognizable, but also made the faith that has been built for decades collapse.
What was not cared about at that time has now become a short board that everyone needs to make up.
2
In 2008, Huawei HiSilicon's mobile phone chip was almost sold along with the mobile phone. At that time, Ren Zhengfei felt that the mobile phone was not the main track of China, if it were not for the financial crisis, the buyer's willingness to acquire was greatly reduced, and there would be no later Huawei HiSilicon Kirin chip.
Huawei's mobile phone was mainly a carrier market at that time, the quality was not good, the reputation was not good, and the insiders did not use it.
However, Huawei set up a project to make mobile phone baseband chips very early, dating back to 2003, five years later to launch the first mobile phone chip, when it was not called Kirin, but called K3. K3 is a peak in the Himalayas at an altitude of 8051 meters, because it is the first generation of products, it is called K3V1.
The result is not as good as everyone thinks. At that time, due to cost considerations, K3V1 and later K3V2 used a GPU (image processing unit) of a small American business, which was cheap but not very compatible. As a result, other big manufacturers do not want it, and Huawei does not use it internally.
Finally, it had to be sold to a copycat mobile phone manufacturer, and it was in this way that K3V1 achieved million-level shipments. But the cottage machine did not make HiSilicon make money, and also smashed the word of mouth. Customers are scattered, the price is cheap, and a lot of adaptation work needs to be done, and the chip itself is backward in process, using the 110nm process, when the mainstream chips have used 65nm or even 45nm.
Huawei terminal company that has done a guinea pig, the inner HiSilicon chip at this time is resistant. If you use Qualcomm chips, Huawei terminals can be done by only two people, and mediatek chips require three people, while eight people with HiSilicon chips can't be done.
When the Huawei P6 mobile phone project was established, whether to use hiSilicon K3V2 was very controversial internally. But at that time, profound changes had taken place within Huawei. The consumer business has become one of Huawei's three major business segments, Yu Chengdong is a leader, and Yu Chengdong, who has worked hard from Huawei's wireless business, knows the importance of chips.
Supporting their own chips is still very beneficial - it can reduce costs, but also differentiate, and can also be used as a marketing selling point. For example, "at that time, HiSilicon's set was about 600 yuan cheaper than Qualcomm", which was a huge driving force for Huawei, which was quite sensitive to profits.
Yu Chengdong insisted on and specified that he would use the K3V2 developed by HiSilicon on Huawei's high-end mobile phones. Later, the Kirin series replaced the K3 series, and the real word of mouth was the Kirin 920. To the Kirin 950, HiSilicon can basically achieve the same level as the industry's leading level.
From 28nm to 16nm, to 10nm, to the latest 5nm, from the perspective of recent generations of technology, HiSilicon has been the first customer of each advanced process of TSMC. Because consumers can directly perceive, The Kirin chip has brought a word-of-mouth effect to HiSilicon.
Many people think that the Kirin chip has made Huawei mobile phones, but it is only half right, which is actually mutual achievement. Without Huawei's mobile phone as a guinea pig, without huge shipments, it is impossible to make a chip continue to evolve iteratively. However, the performance of the follow-up Kirin chip is also a strong foundation for Huawei mobile phones to call the wind and rain in the high-end market.
Huawei can do it, in addition to the determination and strategic vision of the top level, it is also related to historical precipitation. Huawei, which started as a communication device, has long developed communication equipment chips, and mobile phone chips essentially require an understanding of communication networks. Despite the differences, some of the abilities of the two are universal.
But Huawei's highlights on mobile phone chips highlight the loss of another domestic communications giant, ZTE, on mobile phone chips.
As the same city derby, the growth paths of the two companies are relatively similar, both are mainly engaged in telecommunications equipment, both have their own network chips, and the team size is very large.
ZTE made chips very early, starting almost at the same time as Huawei. At that time, the purpose of everyone's chip was also relatively pure, and there was no sense of home and country, more for cost saving considerations. Telecommunications equipment seven countries and eight systems, not their own research and development of chips is not competitive at all.
Many people don't know that ZTE actually has its own mobile phone chip. As early as 2003, ZTE established a wholly-owned subsidiary, ZTE Microelectronics Technology Co., Ltd., on the basis of the IC design department, and the primary task was to develop 3G core chips. Two years later, ZTE developed and mass-produced the first WCDMA baseband treatment kit.
In 2014, ZTE launched the 4G baseband chip "Velociraptor 7510", which is known as the first 4G baseband processing chip based on the 28nm process in China.
But even with so many halo blessings, ZTE's mobile phone chips still do not have much sense of existence. In 2018, Apple announced that it would no longer use the baseband chip of the old man Intel. ZTE also left a message on Weibo: In fact, Apple can consider our Velociraptor 1020 BPU. However, due to the impact of the US ban, ZTE's Xunlong core has disappeared since then, and now ZTE mobile phones use Qualcomm or MediaTek chips.
At one time, some people lamented on the Internet, why can Huawei's Kirin chip succeed, while ZTE's Velocron chip has no sense of existence?
ZTE is not without opportunities. In 2014, ZTE once hoped to promote the chip with the carrier market, and planned to launch about 30 4G mobile phones. As a result, China Mobile changed its strategy for customizing mobile phones, requiring all customized phones to support five modes. ZTE's chips only supported four modes at that time, and ZTE had to postpone the progress, and the shipment of 10 million units was adjusted to 5 million units.
In addition, the chip iteration speed of smart phones is much faster than that of functional machines, which also increases the difficulty and cost of research and development. The chip not only tests the technology, but also pays great attention to the scale effect, if it is not Huawei's mobile phone shipments, the Kirin chip may not be able to do it.
ZTE was mainly based on thousands of yuan machines, and there were few external customers, and it could not bear this investment at all by relying on its own sales. Hou weigui has put forward three requirements: one is that the smart machine chip will be done sooner or later, the second is to do the highest gold content, and the third is to do the part that ZTE is best at. Combining the three, ZTE found that it is more consistent with the communication chip.
In 2021, ZTE terminal shipments will exceed 100 million, of which 50% will use self-developed chips, but most of them are some IOT devices and some home gateway devices, and mobile phones are rarely seen.
3
In October 2014, Xiaomi quietly established a chip company, and Xiaomi had only been in business for four or five years. Xiaomi, Hammer and other Internet mobile phone brands are all software start-ups, making an operating system based on the Android open source platform.
No one expected that Xiaomi would put a big move and directly pull the difficulty coefficient full. Boss Lei later explained, "Because the chip is the commanding heights of mobile phone technology, Xiaomi must master the core technology to become a great company." "The translation is that to become Rebus, you have to make chips.
In fact, these reasons are relatively virtual, and another important reason is that Xiaomi had money at that time. In 2014, the sales of Xiaomi mobile phones exceeded 60 million units, an increase of 227% year-on-year. Capital also favored Xiaomi, giving a valuation of $45 billion, becoming the second most entrepreneurial unicorn after Ant Financial.
Under such a good situation, Xiaomi once thought that it had the ability to make chips, "starting at 1 billion yuan, spending 1 billion US dollars, and producing results in 10 years."
Xiaomi's founding team has not engaged in chips, most of them are Motorola people, and they certainly can't do it on their own. That's okay, just spend money on it. Xiaomi spent 100 million yuan to buy the technical solution of Datang Telecom's Lianxin, and also dug up some people to come over.
At that time, it also bought a license for the chip architecture from Arm. According to the money that Xiaomi paid to Arm, Lei Jun really wants to do a big job.
If you have money alone, Xiaomi has also poached Zhu Shangzu, a technology bull at MediaTek. When he was still in MediaTek, Zhu Shangzu painstakingly persuaded mobile phone companies not to develop their own chips, everyone use MediaTek's on the line, hello, I am good, "as long as mobile phone companies understand some mathematics, they will not develop their own chips, because it is not economical." ”
As a result, the ass decides the head. Within two years, Zhu Shangzu left MediaTek and joined Xiaomi to develop chips.
Xiaomi's core is the same as making mobile phones, adhering to the Internet set and maintaining a consistently fast. 9 months of film, 28 months of mass production, people in the industry have listened to scalp tingling, feel that Xiaomi does not respect the law of chip research and development, so it is certainly not OK. At that time, Wang Xiang, the chairman of Qualcomm China, had not yet joined Xiaomi, otherwise he should also advise Lei Jun not to be so impulsive.
At the beginning of 2017, Xiaomi's first main chip, the surging S1, was released. Xiaomi held a press conference at the National Convention Center, announcing that it will become the fourth mobile phone manufacturer with chip self-research capabilities after Apple, Samsung and Huawei.
But Lei Jun still underestimated the difficulty of the chip.
Although the surging S1 was released to the public, the performance was average. Only the Xiaomi 5C became a guinea pig, and the product performance and market reputation were not good. This also makes Xiaomi a little hesitant to make cores. At that time, it was also rumored that Nokia's new mobile phone would use this chip, but it did not work.
Xiaomi, which is about to go public, cannot endure the pressure of chip investment for a long time. Enterprises want to go public, the financial report can not withstand such a toss, and it is expected to suspend development.
After the surging S2 difficult delivery, the chip master Zhu Shangzu also ran to OPPO, and the news of layoffs and departures continued to come out. People once thought that Xiaomi would completely abandon the research and development of mobile phone chips. But Xiaomi did not give up, but launched an ISP chip, which is a module in the system-level chip responsible for processing image data, reducing the difficulty and no longer hoping to eat a big fat man.
Vivo thinks the same way.
A few years ago, the outside world found that vivo was also recruiting chip talents, and they all thought that vivo wanted to engage in chip research and development with great fanfare. But vivo's ambitions are not so big, more to understand consumers, in order to have the ability to understand the underlying technology.
In the past, after chip manufacturers completed the tape-out, they would cooperate with mobile phone manufacturers to develop adaptations, and for a head company, such a process could not afford to wait. It needs to step in earlier, which also requires a sizable number of chip teams.
Vivo and Samsung have jointly developed a dual-mode 5G SoC, and Vivo has invested more than 500 professional R&D engineers before and after, which lasted 10 months and added more than 400 functional features accumulated to the Samsung platform.
The so-called joint research and development is basically the research and development of chip factories, and mobile phone factories raise demand and participate in evaluation. There is also a reason for this, mobile phone innovation has entered a bottleneck period, and everyone is looking for differentiation. The risk of making chips is too difficult and too high, but the ability to understand chips must be available in order to better match the needs of upstream manufacturers. Now vivo has also released a self-developed IPS chip, and has also begun to practice in the novice village.
As Hu Baishan, executive vice president of Vivo, admitted: "SoC investment is very large and it is difficult to bring differentiated advantages, and there are already Qualcomm, Samsung and other companies in the industry, vivo will not intervene for the time being." ”
But OPPO doesn't think so.
OPPO is more ambitious. Founder Chen Mingyong said that 50 billion yuan will be invested in 3 years. After hearing the news, the TV station that was waiting to be fed was cold and cold, because it was necessary to invest more in research and development, OPPO stopped the variety show, began to search for chip research and development talents everywhere, formed a chip team of thousands of people, and the number of patents was getting more and more every year.
But OPPO has always been very cautious about the outside world, one is a big trick; the other is also related to the style of doing things in this branch, before it is made, it will not say too much, saving embarrassment in the face.
In the past, many people felt that OPPO and vivo had a deep historical relationship, but the two companies were not as close as everyone thought. Chen Mingyong will not look for Vivo as a family member Shen Wei once a year, private competition is fierce, one does, the other will quickly imitate and follow up. It's just that in the early days, some rules were set, such as not robbing each other of talents, going to one, and not going to another in the future.
In the matter of core making, the mentality and practice of the two companies began to diverge. Vivo still insists on following the strategy, dare to eat meat for the world, apple Samsung eat meat, I follow the drink of soup. The road of self-developed chips is still left to others, not within my ability.
Nowadays, in the mainstream mobile phone manufacturers, Huawei's 5G chips are suppressed and have to sell 4G mobile phones; the glory of the split is not said to have self-developed chips for the time being. Vivo and Xiaomi are also making cores, but the overall action is not large, only OPPO rushed to the front line.
As its president of China said, "OPPO has never engaged in real estate, we have to invest money in a longer business, analyze and analyze, put money into chips, the business will be longer."
For a long time, Huawei can succeed, not only with determination and ambition, but also because of decades of technology accumulation in the field of communications, while OPPO has not accumulated in the field of chips and communications, and the difficulty is really like the trench of Mariana, which is bottomless.
No one knows the outcome in advance. However, after 20 years of development of domestic mobile phones, the pattern has been generally stable, and everyone is no longer keen on marketing and price wars, but has a higher pursuit and pursues the real core things. In the past, everyone was unconscious about the chip, but now it has become active.
On the surface, the industry is not so lively, but in fact it has a more difficult task and mission. We bless them.
This article is original content for Digital Intelligence Frontline (szqx1991).