Author: Wen Yu, Editor: Xiao ShiMei
"Foreign auto factories have begun to invest in TSMC for production capacity, and our auto factories will only call there."
At the recently held 2022 China Electric Vehicle 100 People Conference, the former leaders of the Ministry of Industry and Information Technology showed no mercy in their speeches, and the spearhead was directed at domestic car companies.
Behind this beating, China's electric vehicles have reached a crossroads of destiny.
【Missing core】
The real reason why the automotive industry has begun to discuss the core with great fanfare is that it has been plagued by the "lack of core", which is not only a problem for China to face, but a common problem for global car companies.
According to data previously released by AutoForecastSolutions, chip shortages have led to a cumulative reduction in global car production by about 10.2 million units last year. Some of the cars put on the market belong to the "missing core" version, that is, the first car to postpone the replacement of the chip.
The crux of the matter is that until now, instead of being mitigated, the problem has intensified.
In February, Great Wall Euler announced that black cats, white cats and other models will no longer accept orders, and haval H6 will have to wait more than a month. At about the same time, Volkswagen CEO Herbert Diess made it clear that volkswagen would continue to reduce production at overseas factories this year in response to chip shortages. Toyota lowered its domestic production target for April-June by more than 20 percent after further reducing production in March.
If you want to get the chip, it is not impossible, you have to add money.
On March 24, STMicroelectronics took the lead in announcing price increases across Q2, including in-transit and backlogs. Other friends followed up, and a new round of price increases was already on the way.
In April, Ford announced that its Flat Rock assembly plant, which uses to produce Mustang cars, was suspended, while GM also gave it a two-week holiday at a pickup truck assembly plant in Indiana. In short, everything is moving in a worse direction.
In contrast, in the face of this round of lack of cores, Tesla can not do its own thing, but it does not seem to hurt the bones.
According to the data, in 2021, Tesla's global sales reached 936,000 units, a year-on-year increase of 87.4%, and the main models Model 3 and Model Y both achieved a jump in delivery last year.
In 2021, BYD's annual automobile sales reached 730,000 units, an increase of 75.4% year-on-year, and the sales of new energy passenger cars were 594,000 units, a year-on-year increase of 231.6%.
Tesla and BYD can reduce damage in large part due to the self-development of the chip. In particular, BYD, the company entered the chip in 2002, and at present, BYD Semiconductor has become the world's second largest in the field of IGBT after Infineon, not only self-supplied, but also exported.
In order to get rid of the shortage of chips, domestic car companies have devoted themselves to chip research and development since last year. But from a realistic point of view, this move has no effect on solving the chip shortage and stabilizing the supply chain, because the chip development cycle takes about 5 years, and even if the research and development is successful, the crisis of missing cores has passed.
What car companies really want is to seize the commanding heights of the entire industry discourse power and core interests in the future.
【Lesson】
Ten years from now, how much profit can OEMs get from a new energy vehicle?
Before answering this question, let's take a look at PCs and smartphones.
PC jianghu, Microsoft and Intel formed a terrifying "Wintel" alliance, while firmly controlling the speed and direction of the entire industry iteration and development, while eating, the most oily layer of soup into their own bowls. So that to this day, a 5,000 yuan computer, Lenovo can only earn more than 100 yuan, and Intel can easily roll up thousands of yuan of profits, Microsoft is sitting on the ground to collect rent.
In the field of smart phones, Chinese mobile phone brands occupy half of the global smartphone market, but the proportion of profits is only 12%. In contrast, Apple, which has a market share of less than 20%, takes 75% of the profits of the global mobile phone market.
The difference between domestic mobile phone factories and Apple is that the former pursues "take-ism", and there is no difference in configuration to use Snapdragon + Android, and the ugly point is to "work" for Google and Qualcomm; while the latter is self-reliant, self-developed A series chips + IOS systems.
In the entire Apple industry chain, the hard work of Chinese enterprises is the greatest, but the credit goes to Apple. The most intuitive evidence is that more than 90% of the more than 200 million mobile phones apples sold around the world every year are produced in China, but Apple makes money in a year equivalent to dozens of times the profits of the top ten largest A-share Apple concept stocks combined.
The above data reveals a basic fact of the current global industrial division of labor:
Who can get stuck in the most core link of the industrial chain, who can get more cakes, and the links with low technical gold content can only eat some leftovers, and even become sweatshops.
The future of electric vehicles is the same fate.
So what is the most core link in electric vehicles?
From fuel vehicles to electric vehicles, the biggest change is not that energy has changed from oil to batteries, but that the original mechanical industrial product has become a digital product. The core components of fuel vehicles are the three major parts such as engines, gearboxes and chassis, and the future control of the lifeblood of intelligent electric vehicles will become chips and systems.
The new energy vehicle industry has long had the saying of "software-defined car", from automatic driving, intelligent cockpit, to the intelligence of the body and chassis, to the vehicle-road collaboration, the Internet of Vehicles, the operation of intelligent electric vehicles does not need the integration of software and hardware all the time, and even the public that has been pursuing hardware innovation has openly admitted that 90% of the innovation of future cars is gathered in software.
At the same time, the realization of intelligence and electrification requires powerful performance and huge volume of chips to provide support.
There are about 500 chips on traditional fuel vehicles, but highly intelligent electric vehicles may require about 2,000 chips. Tianfeng Securities pointed out in the latest research report that by 2030, the proportion of automotive chips in the total cost of automobiles will reach 50%.
The problem now is that China has no presence in the automotive chip industry.
Data show that the current scale of the domestic independent automobile chip industry is less than 15 billion yuan, the overall localization rate of automobile chips is less than 5%, and the field is basically monopolized by overseas manufacturers. Taking the MCU, the most sought-after master chip, for example, Renesas Electronics, NXP, STMicroelectronics, microchip and other companies collectively control more than half of China's market.
Only in terms of the current situation, Tesla has become apple-like, and the fate trajectory of domestic new energy vehicle companies is more and more like that of domestic mobile phone manufacturers.
The consequence is that not only will the lucrative profits of the entire industry be surrendered, but also the right to speak will be completely lost.
As the world's leading self-driving chip, Mobileye at its peak controlled more than 70% of the market in this field, which can be described as the same in front of car companies.
In order to better adapt to the needs of the terminal, Tesla has proposed to develop its own self-driving technology on Mobileye's chips and algorithms, but the latter has rejected it. Because of this, Musk has strengthened his belief in self-developed chips.
The new domestic car-making forces are also plagued by Mobileye.
In 2019, because Mobileye refused to share the collected data with Ideal, Ideal One equipped with Mobileye EyeQ4 chip could only install another data acquisition camera on its own. Because Mobileye's chip does not support car companies to modify the algorithm themselves, WEIlai can only write its self-developed algorithm into another chip when developing NIO Pilot Pilot Assistance, and the result is that the operation of the system becomes more complex and inefficient.
Not being able to modify the algorithm means that it can only accept products that are consistent with the market, and then lose the possibility of differentiated competition, and eventually plunge into the dead end of homogeneous competition and be forced to accept the fate of low-end inner volume.
In this way, China's new energy vehicle industry has reached a new crossroads, just like the former ministry of industry and information technology leaders mentioned at the 2022 China Electric Vehicle 100 People's Conference, the country played well in the first half of the new energy vehicle competition, but the second half decided the winner or loser.
So does China still have a chance to catch up in the field of automotive chips?
The answer is hope.
【Way out】
Car companies create cores, and the future can be described as mixed.
"Happy" is that most of the car specification chips belong to mature processes, and to a large extent, they are not subject to Moore's Law, which is equivalent to leaving a back door for China.
For RF, power management, MEMS, CMOS sensors and other chips, although the size, speed and power and other indicators are also important, it is not necessary to integrate functions on the most advanced nodes, but to integrate performance, integration and cost, so the absolute advancement of the process is not an absolute truth in the field of automotive chips.
Infineon, for example, launched the 7th generation of IGBT products in 2018, with a 25% reduction in area compared to the 4th generation products, but until now, the 4th generation product is still the mainstream of the market, and the product was released by Infineon in 2007. That is to say, even if the technology of Chinese companies is more than ten years behind in the car rules chip, there is still hope of catching up.
"Worrying" is that the car specification chip is still a part of the chip industry after all, and it also has extremely high requirements for technology, funds and talents.
Musk had plans to develop his own chip in 2013, but he had to cooperate with Mobileye in the early days, and later poached the "chip god" Jim Keller from AMD to have the FSD chip.
Even if talent and R&D funds are in place, if you want to continue to iterate, you must also turn over the mountain of scale efficiency.
According to the industry's previous calculations, in order to maintain the continuous iterative research and development of a car SoC, the annual shipment volume must reach the level of one million. If there is no large enough amount to support, it is impossible to dilute the early R& D costs through commercialization in the later stage, and it is impossible to continue the virtuous circle, unless the outside continues to throw money.
In 2021, Tesla sold 936,000 vehicles worldwide, which is customary to basically have the strength of chip self-iteration. For most of the new car-making forces in China, there is obviously no such ability at present. According to the data, last year, Weilai, Xiaopeng and Ideal delivered 91,429 vehicles, 98,155 vehicles and 90,491 vehicles respectively, and have not yet crossed the 100,000-vehicle mark.
Since last year, the domestic automotive industry's "core-making" wave is higher than a wave, radicals such as zero-running cars, once shouted out "do not make chips of car companies are not a good technology company" rhetoric, resulting in many car companies to follow the trend, began to build their own team to organize chip research and development. However, from the perspective of reality, this kind of fragmented approach will not only cause waste of resources, but also make it difficult to be optimistic about the results.
It is better to concentrate on doing big things and support the existing domestic chip forces through shareholding and joint ventures, similar to power batteries. For car companies, they not only retain the right to participate in research and development, but also lock in part of the future profits in advance; for chip companies, they have both received financial support and orders, and can give full play to their experience and advantages.
In fact, some car companies also do this.
In 2021, Ideal held hands with Horizon, Dongfeng held shares in Huaxin, Geely joined hands with Core Energy and Core Technology, and in March this year, Xiaopeng and FAW made investments in Shanghai Zhanxin and Core Qing Semiconductor Company respectively. Unconsciously, the tentacles of car companies have been widely extended to chips.
From a historical perspective, semiconductors have never been an industry that can exist and develop independently, and the upstream and downstream must be highly coordinated, and equipment, raw materials, manufacturing capabilities, and markets are indispensable. The future of automotive chips is by no means a solution that can be solved as long as car companies participate, but the joint efforts of the entire domestic chip industry.
Ye Tianchun (chairman of the Integrated Circuit Branch of the China Semiconductor Industry Association) mentioned a data in a previous speech that is worth noting, from 2016 to 2020, the proportion of domestic companies in the top ten wafer manufacturing companies in Chinese mainland has dropped from 44% to 27.7%, but TSMC, Intel, Samsung and other companies are constantly expanding in the Chinese mainland. To a large extent, this means that the domestic replacement of Chinese chips is not advancing or retreating.
For car companies, core manufacturing is the only way out without a way out, but this road is really difficult to walk.
disclaimer
This article involves the content of listed companies, which is the author's personal analysis and judgment based on the information publicly disclosed by listed companies in accordance with their statutory obligations (including but not limited to temporary announcements, periodic reports and official interactive platforms, etc.); the information or opinions in this article do not constitute any investment or other business advice, and Market Value Watch does not assume any responsibility for any action arising from the adoption of this article.
——END——