There is no future for cross-border "core building" of car companies

No matter how bright the future of new energy vehicles is, it is difficult to illuminate the future of cross-border core manufacturing of car companies.

”

The author | Bao Yonggang

Editor| Wang Yafeng

The old guns in the chip circle do not seem to be optimistic about the self-developed chips of car companies.

This is obviously in stark contrast to the growing enthusiasm of car companies for making cores and the enthusiasm of capital.

The keepers are pessimistic, the crossovers are crazy, the pessimists are not necessarily correct, and the optimists cannot touch the stones and cross the river.

"If auto OEMs choose to develop chips completely, I think most of them will end in failure." Many industry insiders have told Leifeng Network: "Even if the automotive OEMs split the chip team and develop independently, the probability of success is not high." ”

In the past few years, there have been tesla, Weilai, Xiaopeng and other new car-making forces, and then many traditional car companies such as Volkswagen, Hyundai, geely, etc., have announced their plans for self-developed chips.

"The biggest driving force for cross-border core manufacturing of new force car companies is to tell stories to the capital market, new force car companies have no experience in self-developed chips, the path is still relatively radical, and it is difficult for traditional car companies to apply self-developed chips on a large scale." There are many unpromising voices in the industry.

Putting aside the voice of doubt, some people believe that the lack of cores that began in 2020 is an important driving force for automotive OEMs to develop their own chips. Some people also believe that the core appeal of auto companies' self-developed chips is to break the shackles of traditional automotive chips, meet the needs of differentiation, and enhance competitiveness.

In any case, the automotive OEMs are scrambling to gain more voice through self-developed chips, which indicates that a change is brewing in the automotive industry chain.

1

Is it a self-developed chip or a capital game?

Chip investors invested in Jiuding said bluntly: "Automotive OEMs' self-developed chips will increase our willingness to invest." But it is also necessary to judge whether it is really a research and development strength rather than a concept, and whether the strategy and planning are reasonable. ”

Capital is indeed a necessary condition for the new car-making forces to develop their own chips, after all, a high-performance autonomous driving chip investment is calculated in hundreds of millions of dollars.

A typical representative of cross-border self-developed chips of auto OEMs is Tesla, whose first self-developed chip FSD (Full Self-Driving) was released in 2019 and was the first to be installed in the Model 3.

Seeing that Tesla has used self-developed chips, the new domestic car-making forces have also acted quickly. In 2020 alone, players such as Weilai, Zero Run, and Xiaopeng have successively broken out rumors about core building. Among them, zero-run cars also said that "car companies that do not make chips are not a good technology company."

In less than two years, almost all the new car-making forces have embarked on the road of self-developed chips.

"Self-developed chips can allow new car-making forces to give the capital market new imagination space and obtain additional benefits in the capital market." After Tesla and some new domestic car-making forces announced their self-developed chips, the stock price rose 3-4 times in one year. Zhang San, who has more than 20 years of experience in the automotive industry, pointed out.

In addition to attracting capital, many people believe that this wave of automotive OEMs' self-developed chips boom is due to the continuous shortage of automotive chips.

A number of chip industry experts have said that the cycle of self-developed chips is very long, far from quenching the near thirst, and the lack of cores is a factor for automotive OEMs to decide on self-developed chips, but it is not the key.

Manager Wang of the Power Electronics Division of Shanghai Fudan Microelectronics Group believes: "The self-developed chips of automotive OEMs help them better grasp the chip industry chain, and keep the supply chain in a small fluctuation range when the core is missing." But more critically, auto OEMs want to gain more right to know and have the right to speak. ”

Xinmou Auto analysts also said, "The oem's self-developed chip itself is to compete with Tier1 (a first-tier supplier) for the future of automotive architecture development and the right to speak in the industry." ”

New car-making forces and traditional car companies have different needs for higher discourse power through self-developed chips.

2

The new car-making forces have developed their own chips with radical targets

The new car-making forces represented by Tesla have almost given priority to the high-performance AI chips needed for autonomous driving when developing chips themselves, and their goal is to achieve differentiation, but this is a bold and even radical approach.

"The style of the new car-making forces is more radical as a whole. Their aggressive style has previously been recognized by some capital and the market, so it is understandable that they dare to challenge the difficult high-performance AI chips at the beginning. Xin Shi, a veteran in the upstream of the automotive industry chain, believes.

Xu Chao, vice president of Xinchi Technology, analyzed from the value orientation of the new forces of car manufacturing, "This is related to the path of the future evolution of the new forces of car manufacturing products, in the field of automatic driving, they have more room for innovation, and this value orientation makes them choose chips related to self-developed automatic driving functions." Moreover, because autonomous driving and algorithms are closely related, the implementation of the data and algorithms accumulated by itself can be more optimized through self-developed chips, which has more advantages and barriers than the use of general-purpose chips. ”

Xin Shi said, "Automakers also hope to avoid the PC industry's whole machine factory being reduced to Intel and Microsoft part-time workers in the automotive industry again." The whole machine factory makes its own chips and algorithms, which can ensure the stability of the supply chain and pricing power, and can launch products with their own characteristics."

How important is differentiation? The automotive industry has a century of history, and today the wave of electrification and intelligence is unstoppable. Under the new trend, intelligent experience has become an important decision-making factor for young people to purchase new energy vehicles, and has also become a key path for major car-making players to break through in the fierce competition.

However, the existing automotive industry chain will limit the development of the automotive industry in a smarter direction to a certain extent. For example, in accordance with the traditional way, the center console screen of the automobile OEM can only be based on the solution provided by Tier1, and the size and resolution of the screen are very small, and it is difficult to differentiate.

Intelligent cockpit and automatic driving functions are also an important embodiment of differentiation, and automatic driving is closely related to data and algorithms, which are the unique advantages of automotive OEMs, which are reluctant to share with chip companies.

"System manufacturers are increasingly reluctant to share their unique needs and understanding of application scenarios directly with traditional chip design companies, but form self-developed chips through various forms, and combine systems and software to create an exclusive ecology, which is a trend I see." Yao Ye of the EDA industry said.

Chapter 3 pointed out, "Tesla is a special case, and it cannot be used as a special case as the development trend of the industry. Tesla is the first in the new energy vehicle market, its product premium ability, marketing ability, and reserves in the chip field, which are all capabilities that other new car-making forces do not have. Tesla felt the stones to cross the river, and other new car-making forces touched Tesla to cross the river, and the risk was completely different. ”

3

Traditional cars seek stability, and it is more difficult to "break"

Compared with the new forces of car manufacturing, traditional car companies in the opening of self-developed chips when the choice is more conservative, many car companies have chosen this round of core shortage of the most serious MCU, although compared to the high-performance automatic driving chip is less difficult, but there are enough MCU to choose from on the market, car companies can be fully applied in their own products, affecting the success or failure of traditional car companies to make cores.

Analysts at Xinmou Automobile pointed out that "the products of the new car-making forces are mostly new energy vehicles and hybrid vehicles, while traditional car companies have both internal combustion engines and new energy vehicles." The product layout has led to the short-term necessity of traditional car companies to be more inclined to ensure the safety of the supply chain of two different power types at the same time. ”

"Compared to other chips, MCUs are like brains, with long replacement cycles and high investment." Manager Wang said, "The most serious core-missing core in this round of cars is the MCU, although the unit price is not high, but it is indispensable for the whole vehicle." ”

Market data can also explain that traditional car companies' self-developed MCUs are a better choice. Strategy Analytics data shows that in traditional fuel vehicles, MCU value accounted for the highest proportion of 23%, in pure electric vehicle models, MCU value accounted for the second, at 11%.

Source: Strategy Analytics, Gaz Automotive, Guohai Securities Research Institute

Of course, MCUs are also divided into high-end and mid-level. Manager Wang explained, "The requirements for computing power and advanced processes in the low-end car regulation MCU are relatively less high, and the scenes that can be covered are more extensive. High-end MCUs are not only more technically demanding, but also relatively more dedicated. The soC chips developed by car companies are more valuable. “

Zhang San believes that "there are many MCUs to choose from in the market, and the cost of replacing mid-to-high-end MCUs in existing products is relatively high, and the MCU developed by traditional car companies is expected to be difficult to apply in their own products on a large scale." ”

4

The losers will be the majority

"Any market is bound to start with a hundred flowers, and in the end there will only be a few left." Even if this result can be predicted, it still cannot prevent automotive OEMs from entering this market. Xin Shi pointed out, "As OEMs begin to develop their own chips, the pattern of the automotive industry chain is also changing." ”

The biggest advantage of automotive OEMs' self-developed chips lies in the understanding of application scenarios, no worries about no customers, and the process of chips from design to verification is faster.

The disadvantages of automotive OEMs self-developed chips are also very obvious, due to the lack of talents in the entire automotive chip industry, it is a challenge to be able to recruit suitable talents, and to design suitable chips under the system of automobile companies.

"Compared with designing a suitable chip, ensuring the continuous and stable operation of the chip throughout the life cycle is a greater challenge for auto companies to develop their own chips." Manager Wang said: "Doing chips is not achieved overnight, this experience depends on time accumulation and application polishing, even if the main engine factory has enough experience authority in terms of the working conditions of the vehicle chip and the test standard of the vehicle system, but really design a chip to verify the self-developed mass production of a chip that meets the high reliability standards of the car, it is also a new challenge for the main engine factory There is still a long way to go." ”

Xu Chao further said, "There are not many talents in the world who have successfully designed and mass-produced automotive chips, and to ensure the safety and reliability of chips, it is necessary to have more than ten years or even decades of experience in the industry, and half of the accumulation of experience is a lesson." ”

Many people in the industry have expressed such concerns, on the one hand, the global automotive chip talent is very small, if a chip software (or hardware) engineer to be a product manager, the results will certainly not be good. On the other hand, automotive OEMs may not have a deep understanding of the chip industry, too optimistic, it is difficult to define a chip that meets the actual application needs, increasing the probability of self-developed chip failure.

Xin Shi also pointed out that the challenges at the software and ecological levels are greater than those at the chip hardware level. This is because to play the hardware performance advantages of self-developed chips, there must be a better software and hardware integration. Developing a working software stack takes much longer than designing a chip.

There is also a prominent challenge in the self-developed chips of car companies, that is, the customer is single, if there is not enough shipments, it is difficult to recover the cost, and it is even more extravagant to reduce the cost of the chip of the whole vehicle.

Leifeng network learned that according to the different operating efficiency of each company, the annual shipment of an automobile SoC is less than close to one million and more than one million, which is difficult to support the continuous research and development investment of chips. Tesla, the global new energy vehicle sales champion in 2021, delivered 936,200 vehicles. The sales volume of the first three wei Xiaoli of the domestic new forces has not exceeded 100,000 vehicles, and its own hematopoietic ability needs to be improved.

However, with the development of smart cars, the market for automotive chips is also growing rapidly. Intel CEO Pat Gelsinger predicts that by 2030, chips will account for more than 20% of the bill of materials (BOM) of high-end cars, a five-fold increase from 4% in 2019. In addition, by 2030, the overall market size of automotive chips will more than double to $115 billion, accounting for about 11% of the entire chip market.

Based on the advantages and disadvantages of auto main engine factories' self-developed chips, Zhang San said, "I think that only by having relatively good vehicle sales in the world, as well as giant car companies with strong innovation capabilities and resources, and with certain chip technology accumulation and talents, it is possible to succeed in self-developed chips." ”

A number of industry experts also pointed out that "whether it is a new car-making force or a traditional car company, even if the chip team is split into independent operations, it is difficult for the automotive OEMs to finally succeed in developing their own chips." Due to the consideration of technology leadership, price, system adaptation, etc., it is still difficult for the split team to obtain orders from other customers, including competitors, and it is difficult to increase chip sales. ”

Xu Chao believes that automotive OEMs can choose to cooperate with neutral chip companies. This method is not a new model in the field of automotive chips, but also an option for the oem's self-developed chips, and it is a better choice. The two sides have in-depth cooperation, based on the mature architecture platform of automotive chip enterprises, to a certain extent of customization of the special needs of car companies, such as 5%-15% of innovative functions. This can not only ensure the differentiated needs of car companies, but also cooperate with experienced teams on the basic chip architecture of mature mass production, which can greatly reduce the risk of R&D failure.

In either way, self-research shows that under the new development trend of the automotive industry, there will be closer cooperation and communication between automotive OEMs and chip companies. As for whether this change will significantly weaken Tire1's right to speak, there are positive and negative views, the key lies in the success or failure of the automotive OEM's self-developed chips, and whether Tire1 can actively embrace the change.

Obviously, the automotive OEMs develop their own chips, the demand for differentiation and the higher control of the industrial chain are the core appeals.

As for the outcome, the industry predicts that very few companies can fully develop successfully, most of them will fail, but for the ultimate winner, no one can accurately predict.

END