Are you studying chips for use?
You need to invest 100 billion to research chips.
But if you buy Qualcomm, you only need to spend 50 billion.
Got it?
This is how foreign countries hinder your technological development. When they see that you are going to succeed, they immediately lower the price. Short-sighted people will stop investing and research and just buy, buy, buy.
That’s why I say Huawei is awesome! ! That’s why we need to support Huawei. . . Otherwise, you will still be stuck in the future. .
In fact, according to the domestic atmosphere, not independent research and development is generally the right choice. There is no return and high investment has been sustained for more than ten years. What company would make such a decision? But Huawei has done it, so it’s awesome.
Text/Xiaoyi Ping Technology
There are many reasons why domestic mobile phone manufacturers dare not make chips, which can generally be summed up as follows:
① Funds Consumption, the technical threshold is too high, the research and development cycle is too long, and it requires a lot of time and cost. The company cannot afford it.
②Product positioning problem.
③Restrictions on communication baseband.
④ Pressure from Qualcomm.
As a result, these mobile phone manufacturers are afraid to develop chips, not only before, but also now.
01. The threshold for developing chips is too high, and ordinary mobile phone companies simply cannot afford it.
Chip research and development is a project that requires long-term and heavy investment. First, you need to recruit a large number of talents related to semiconductor circuit design to form a team, and you also need to recruit an experienced boss to be the team leader. The person at the helm of this team controls the overall situation, designs the structure and manages the team.
This kind of talent is very, very scarce. Most of them are well protected by other chip design companies, and their annual salaries are quite high.
Take Jim Keller, known as the "Silicon Immortal" in the industry (he is the founder of Apple's A series, AMD Zen series and Tesla FSD chips). Although his annual salary is not known to the outside world, I know, but according to anecdotal reports, the annual salary is at least tens of millions of dollars. In other words, if you want to support such a person, the salary alone will cost nearly 100 million yuan a year.
Of course, you can also choose to train your own chip design team from scratch, just like Huawei, but this also requires a lot of time and capital costs. And this also requires the company to have a relatively strong technology research and development atmosphere and be able to retain these people. Otherwise, the technical staff will not be retained at all. As soon as you have trained them to some extent, other companies will come and poach them for you. It has to be cultivated from scratch.
Chip R&D itself is an industry that requires long-term accumulation. For a long time, it will be in a situation where only investment has no return. If there are no other business lines as support or the company itself does not have technical genes, it will be difficult. If you persist, the most terrifying thing is that if the R&D team loses confidence, then it’s really over.
Moreover, chip design is a project that extremely tests teamwork, especially for more complex chips. Once the team does not work together enough, the chip will not be designed at all. Because to design a chip, you need to consider a lot of variables, signal interference, heat distribution, the impact on the chip under different manufacturing processes, magnetic field interference, etc. This requires a large number of different teams to work together, and the chip There should not be any flaws in the design. If problems arise, it will have to be completely scrapped and the team will need to be reorganized for research and development. This will consume a lot of resources and have a huge psychological impact on the team.
In addition, when designing chips, you also need to buy licenses for various architectures. For the simplest ARM architecture, do you need to buy a core license? Another investment of tens of millions of dollars. After each chip design is completed, it needs to be checked and simulated through software to detect errors. The cost of each verification is tens of millions of dollars (one dollar per chip). , At present, mobile phone chips require at least tens of millions of gates, everyone calculates the cost by yourself)
Then there is the chip tape-out test. The cost of a tape-out is generally around four to five million yuan, and tape-out can be It doesn't just need to be done once, it needs to be done many times, and it will cost tens of millions to go back and forth.
In short, chip design is still a very complex system engineering that integrates various high-precision technologies. It is really difficult for ordinary companies to do it.
Xiaomi’s CEO Lei once said that chips will be priced at a low price in the future. However, when Xiaomi actually made chips, it discovered that chips are really difficult to make.
02 Product positioning issues
Huawei has entered the smartphone market as a high-end business machine from the beginning. The first-generation Huawei Mate was released in 2013. Its price It was 2,699 yuan, which was higher than the price of domestic smartphones at that time. This was Huawei’s positioning of its own models. In this case, Huawei can and must have more sufficient reasons to create certain differences. , self-developed chips, focusing on information security are a good selling point. Making high-end mobile phones itself requires a relatively long time to prepare, which coincides with the mileage of chip research and development.
On the other hand, other domestic mobile phone brands at that time mainly used cost-effectiveness as the main selling point. For the products of these companies, they only need to ensure the cost-effectiveness and popularity of the mobile phones. There is no need. If they think too much about the so-called exclusive technical advantages, for them, technology will never be as valuable as marketing.
03. Constraints of baseband chips
If the previous dimensions can be solved, baseband is a threshold that is difficult to get around. The reason why many chip design companies are developing mobile chips The main reason behind the failure is the inability to play with the baseband, including the well-known giants such as NVIDIA, Broadcom, and Texas Instruments. They have all launched mobile phone chips, but without exception they all ended in failure.
Because the design of baseband chips not only requires strong chip design capabilities, it also requires design companies to have a lot of practical experience and good cooperative relationships in the communications field, because any baseband needs to operate globally. The manufacturer conducts large-scale compatibility testing.
Moreover, baseband also involves extremely complex patent issues. Many early communication patents were occupied by Qualcomm. If you do not have a communication patent that you can produce, cross-license patents with Qualcomm to reduce the number of patents. If you pay too much, the cost of the baseband chip you develop is very high. As a result, the cost of the chip you build may be higher than buying a Qualcomm chip, and the performance may not be as good as others. You still build a What strength?
If you look at the domestic mobile phone companies, apart from Huawei, only VIVO can be regarded as a company with certain accumulation in communications (it was once a communications company under BBK), and several others Mobile phone brands have almost no experience in communications. In this case, is it possible for you to let them build their own baseband chips? It cannot be made even if it is as strong as an apple.
The baseband used in the ThePaper S1 chip produced by Xiaomi at that time was the baseband of Lianxin. The performance of the baseband was very weak, as shown below, and it was incomparable with the basebands of other companies.
04. Pressure from Qualcomm
Domestic mobile phone brands are now too deeply tied to Qualcomm. It can be said that once Qualcomm gets stuck, these domestic mobile phone brands will be finished immediately. , in this case they have to consider Qualcomm’s attitude, and making chips itself is a challenge to Qualcomm. In this case, do you think Qualcomm will be a good man? Watching the chip you developed fight against him? Don't think about it. Although it is impossible not to sell you chips, it is still easy to beat you. Do you feel uncomfortable if you delay the supply time of chips, reduce the supply quantity, and raise the price? Domestic mobile phone companies can organize to resist, but if you don't organize well, you may be gone. This is the current situation.
Huawei is different. Huawei had no intention of in-depth cooperation with Qualcomm from the beginning when it produced smartphones. Qualcomm initially supplied mid- to low-end chips, and after the rise of Kirin mid-to-low-end chips, You don’t even need mid- to low-end chips, that’s the difference.
In short, it is too difficult for domestic mobile phone companies to develop independent chips, especially since it is impossible for a second Huawei HiSilicon to appear in a short period of time.
It’s just a cover-up, there’s no need to brag about it. Others didn’t think of it before, and the cover-up is considered self-developed.
Everyone knows how high the cost of self-developed chips is.
First of all: researching chips is a bottomless pit that cannot be sustained without certain capital investment. Huawei is the big brother in the communications industry, and Huawei is willing to invest in HiSilicon to make the current Kirin.
Secondly: Even if you enter the chip industry, the cost is not necessarily high. Each tape-out is basically in units of hundreds of millions. If you fail several times, few companies can afford it. After all, the cash flow is now abundant. companies are in the minority.
Finally: R&D takes a long time, and success is not achieved as soon as you enter the industry. After investing investment funds, it may take several years to see the results, or it may disappear directly.
Therefore, everyone should be proud that Huawei can have HiSilicon Kirin, and it is also a mature chip. Of course, companies such as Xiaomi, Oppovivo, etc. have also become bigger and are starting to want to invest in this area. This is a good thing. , but don’t be yin and yang.
Not even now! Make a PPT and hold a press conference, and the stock price will rise. If you don't have time to do the dirty work, it will be nice to use it.
The common mentality of people is "Since it is ready-made, why do you have to do it yourself?"
Let me digress first. I have been working in China for more than six years and what I have seen is generally The phenomenon is: "If you can be lazy, be lazy, if you can't be lazy, be procrastinating." Some customer service information is vague, and they are asked to communicate clearly with the customer. Nine out of ten people do not communicate with the customer, or they say to their face, "I guess it should be..." , or tell you how to do it after a few hours. In the end, most of the parts are wrong and need to be corrected. Some of them are because the customer service did not communicate with the customers, and some of them are customers who are believers in "If you can be lazy, be lazy, if you can't be lazy, then procrastinate." Anyway, what they try their best is to do less. There is no shortage of such talents in the production department. Some materials are of limited use. They will wait for others to make them first, and then pick them up and use them themselves.
Getting back to the topic, the above digression is just to illustrate the mentality of most people, "If you can't do it, try your best not to do it." People can still be lazy when it comes to things as simple as the above. Chips, the most difficult component of mobile phones, are readily available, so why do we have to work so hard on ourselves?
Maybe everyone has heard this kind of logical argument, "This thing is impossible to do. If it is so easy to do, others have already done it, so where is our turn?" This kind of negative spiritual suggestion. There are actually quite a few people like this. They have all decided that they are not good enough, so they will just run their own processing plants peacefully, and they don't even dare to dream about chips.
The last thing is R&D funds and talents. Basically, as an enterprise, not an academic institution, you have to bear the R&D funds yourself. As we all know, chip R&D is a huge black hole of funds. If you only do R&D without profit support, and without talent for R&D, the company will lose money in a few years. Go in.
How many companies can sustain this investment? Currently, in the mobile phone industry, Huawei is the only one in our country
Years ago, there was no big conflict between China and the West. The West spent its main energy on Russia and the Middle East, and they are indeed developed. As long as Chinese universities follow the Majors related to economics must learn Western economics. It can be said that China sincerely wants to learn from the West. Many Chinese entrepreneurs also sincerely believe in the theory that building is worse than buying, and buying is worse than renting. In the past, there was Lenovo, and now there is Xiaomi. Both executives have said similar things. It can be said that they are good students of Western economics. .
A start-up company has limited funds and needs to purchase various raw materials, labor wages, etc. It really does not have much money to invest in research and development, and there are already mature and reliable raw materials on the market. If it invests in research and development by itself, it may be ready as soon as it is developed. If you fall behind, your money will basically be wasted. As a company decision-maker, it would be foolish not to buy directly from others. Ni Guangnan and Ren Zhengfei have both been criticized for this, so it is not difficult to understand the choices of companies such as Lenovo and Xiaomi. One way is to invest 1 billion every year for 5 consecutive years, with a 5% probability of success. If you succeed, you will make a profit of 100 billion. If you fail, you will not get back a cent. The other is to buy it directly without investing, with a 95% probability of success and a profit of 10 billion. If it were you, what would you choose?
Maybe the leaders of these companies have learned Western economics so well that they have forgotten the first lesson, which is that the entire Western economic theory has an assumption that everyone is rational and there is no outside world in the market. Intervention, however, does not exist in reality. However, some business owners still believe in the theory that it is better to buy than to build, cannot recognize the reality, and act as thugs for Western capital. I believe it won't end well.
Of course, most companies are waking up from their dreams around 2018. The United States has also successively provided companies, universities, and individuals with China’s leading science and technology. Now the Chinese people know that it is better to rely on others than to rely on themselves, although it is a bit Late, but better late than never.
Eating meat costs a lot of money, and it is enough to gnaw on other people's bones. Huawei's current situation is that if rich people don't sell you pork, they have to consider raising their own pigs.
The investment in chip talents is very large. Without a good talent incentive mechanism, no matter how rich you are, you will get twice the result with half the effort, and the failure rate is very high. No one doesn’t want to make chips, just look at Motorola, Intel or AMD.