Recently, the automotive chip company Black Sesame Intelligence officially launched a charge to the Hong Kong Stock Exchange. If successful IPO, it will become the domestic automotive chip field of Hong Kong's first shares, and not long ago, another automotive chip company core Wang micro is to the Science and Technology Innovation Board to launch a sprint, look at the country, see can be found throughout the field of automotive chips are booming on the road.
The trend of automotive intelligence is more and more obviously
In the field of new energy vehicles, the trend of electrification, digitization and intelligence has formed a consensus, but the value of "computing" in the new energy vehicle industry has been quite controversial. At present, the value of digitalization and intelligence brought by "computing" in new energy vehicles is becoming more and more prominent, as evidenced by more and more chips on the car, more and more strong arithmetic power, and the value of the proportion is getting bigger and bigger.
According to Frost & Sullivan's report, a new energy vehicle uses more than 1,500 chips on average, which is twice the amount of chips used in traditional fuel vehicles. Moreover, the arithmetic level of the chip on new energy vehicles is rapidly improving, and the value of the chip in the whole vehicle is also gradually increasing, and the chip has now reached 5%-8% of the cost of the whole vehicle.
China is undoubtedly in the leading position in this global automotive industry change. At present, China has become the world's largest and fastest-growing new energy vehicle market, according to car sales, the Chinese market accounted for 64.8% of the world. China is not only a huge market for new energy vehicles, but also has achieved remarkable results in terms of localization, 80% of new energy vehicles sold in China in 2022 will come from domestic automakers. However, with the domestic new energy vehicles a great situation is in stark contrast, most of the domestic new energy vehicles, installed on the chip is not domestic.
China's Automotive Chip Army, Battling NVIDIA
Frost & Sullivan statistics on the global market share in the field of high-computing power (arithmetic power greater than 50Tops) automated driving SoC chips in 2022: in the Chinese market, more than 80% of the automated driving chips come from Nvidia; the domestic Horizon accounted for 6.7%, the Black Sesame Intelligence accounted for 5.2%, and Huawei Hesi accounted for 0.7%, and the three together accounted for 12.6%, totaling approximately 15 percent of the total of Nvidia's chips. Nvidia's 15%.
According to the Black Sesame Intelligence prospectus, its 2022 revenue was $165 million and operating loss was $1.053 billion; net loss attributable to shareholders was $2.7539 billion. As a comparison, NVIDIA's 2022 revenue was $26.97 billion and net profit was $4.368 billion. Of course, automotive chips only account for a small part of NVIDIA's revenue, but its GPUs have a strong applicability in the field of automotive chips, and the huge volume of research and development investment, customer resources, etc. will bring great advantages.
It should be noted that, in addition to Horizon, Black Sesame Intelligence, Huawei Hesse, the three market share is relatively high in the automotive chip suppliers, domestic manufacturers there are Cambrian, Core Toner Intelligence, after the Mo Intelligence, Core Chi Technology and other startups, as well as BYD, Zero Run and other automobile manufacturers self-developed chips, as well as Baidu and other Internet vendors.
In addition, according to the China Automotive Chip Industry Innovation Strategy Alliance Standards Working Group statistics, there are more than 100 domestic enterprises engaged in the development and production of automotive chips, more than 50 chip listed companies claimed to have automotive-grade products or mass production applications. It can be said that although the domestic manufacturers in the market share can not add up to NVIDIA a, but at present or has formed a huge automotive chip system.
Three major obstacles that domestic vehicle chip manufacturers need to confront
For domestic automotive chip manufacturers, to catch up with international giants like Nvidia, the key to success or failure is still to overcome the "three mountains": performance barriers, advanced chip processes and hardware and software ecology. From the table can be found in the arithmetic and process, horizon, black sesame intelligence, huawei heath and other domestic manufacturers and nvidia there is still a huge gap.
A core indicator of automotive chips is the arithmetic power, the greater the arithmetic power represents the stronger its performance, and with the increase in the degree of automotive intelligence, the amount of data that needs to be processed is getting bigger and bigger, and the demand for chip arithmetic power is also increasing at the same time. Higher arithmetic power often means stronger competitiveness. At the same time, the chip manufacturing process is also very critical to improve the arithmetic, the same chip 4nm than 16nm a lot stronger.
In addition to computing performance and chip process, there is a very important but easily overlooked factors, that is, with the chip supporting hardware and software ecology. It can even be said that the construction of hardware and software ecosystem is the biggest competitive barrier for chip manufacturers. When Intel and Microsoft built wintel alliance is a typical example, Microsoft operating system and Intel chip highly adapted to the operating system on top of millions of applications, tens of millions of developers. Chasing Intel's chip performance is relatively easy, but to shake the tens of millions of developers, tens of millions of applications to build the ecosystem, it is as hard as heaven.
In the field of automotive chips, NVIDIA has established a solid competitive barrier through the deep integration of hardware and software and the construction of an ecosystem. Technological innovation is the core, and they rely on the deep accumulation of GPU technology to develop the Drive platform that integrates a variety of AI algorithms to support the various functions of self-driving cars.
In addition, NVIDIA has established close partnerships with major global automakers, parts suppliers, and software developers to form a complete ecosystem, and they have also actively participated in the development of automotive industry-related standards and launched Safety Force Field (SFF), a computational framework for verifying the safety of autonomous driving systems. These strategies have enabled NVIDIA to form a unique competitive advantage in the field of automotive chips. Therefore, whether in terms of chip arithmetic, process or ecological construction, domestic automotive chip manufacturers have a long way to go.
The growth of domestic automakers is not only driving the rise of the domestic chip industry, but more importantly guarding the security of the much larger domestic new energy vehicle industry. China's new energy automobile industry is rising strongly, not only occupies 80% of the domestic market, but also gradually going global, China has now surpassed Japan to become the world's largest auto exporter. According to the China Association of Automobile Manufacturers data show that in the first half of this year, China's new energy vehicle production and sales amounted to 3,788,000 and 3,747,000, up 42.4% and 44.1% respectively, and the export of new energy vehicles 534,000, up 160%.
In this great situation, but there is a huge hidden danger - most of China's new energy vehicles, chips are dependent on imports. The vast majority of Chinese models, its automotive chips, especially high-end chips from Nvidia, Intel, Qualcomm and other foreign giants, which will bring huge risks.
Although the chip accounts for less than 10% of the value of the car, but a car without a chip, the loss is 100%. 2022, the global automotive market has experienced a serious chip shortage. According to estimates, due to the "lack of core" led to the 2022 global automobile production reduction of more than 4 million units, for China's automobile industry, the lack of core brought about by the risk is particularly. Chinese enterprises in addition to facing the market risk brought about by fluctuations in the global industrial chain, but also to face the risk of the chip was "neck".
If China's new energy automobile industry is built on the basis of the supply of U.S. chip companies, in the extreme case, the automobile chip is likely to become a clamp down on China's new energy industry handle. A situation is completely cut off supply, will let China's booming new energy automobile industry instantly into a state of shock. But a more likely scenario is not completely cut off supply, but will limit the chip level.
NVIDIA's high-end chips such as A100 and H100 can no longer be exported to China, and Chinese companies can only get chips that are a generation behind. In addition, Huawei, for example, can obtain Qualcomm chips, but can only obtain Qualcomm 4G chips, not 5G chips, which leads to the 5G era Huawei but can only sell 4G phones.
The degree of automobile intelligence is seriously dependent on the level of arithmetic power of the automobile chip, and in the case of automatic driving, for example, if the chip arithmetic power is not enough, the corresponding automatic driving ability cannot be realized. Assuming a situation, the world's most advanced chip arithmetic power to reach thousands of TOPS, can effectively support L4 above the automatic driving. However, Chinese automobile enterprises can only get automobile chips with arithmetic power below 1,000 TOPS, and their self-driving ability is limited to below L3. In this case, the automotive chip is like the "wise son" of China's automotive industry, "locking" the ability of Chinese car companies.
To break this deadlock, not only do we need the sustained efforts of automotive chip companies like Horizon and Black Sesame Intelligence, but also end-user car companies like BYD, Azure, Ideal, Xiaopeng, BAIC, SAIC, as well as automotive OEMs, sensor makers, and automotive software openers, to join hands to flourish the domestic automotive hardware and software ecosystem.
Happily, some results have been achieved so far. Taking Black Sesame Intelligence as an example, the number of its customers has grown from 33 in 2020 to 45 in 2021, and then to 89 in 2022. It has received intentional orders for 15 models from 10 automotive OEMs and Tier 1 suppliers, as well as cooperation with more than 30 automotive OEMs and Tier 1 suppliers, including giants such as FAW, Dongfeng, JAC, and Baidu.
Horizon's performance is even better, with its Journey series chips having secured the fixed-point cooperation of more than 20 automobile enterprises, 120 models of front-loaded fixed-points, and more than 50 models that have been mass-produced, including terminal automobile enterprises such as Chang'an, Chery, GAC, Ideal, SAIC, BYD, Nezha, etc., and the overall shipment of the Journey series of chips amounted to 2.8 million chips.
