Whether it is Xiaopeng, Huawei, Ideal, Weilai or whom, as long as they are rolling up the city NOA now, they are almost forced out of breath by Tesla.
When He Xiaopeng announced on January 30 that it would push all 243 cities, in 2024, it will compete with Huawei not only in the number of open cities, but even in the number of open communities, and gradually promote China's intelligent driving technology to the world. Everything is not only staring at Yu Chengdong, staring at Huawei, but also pointing the tip of the knife at Musk's Tesla.
The battle for data and enterprise efficiency is a heavyweight point comparable to the Huaihai Campaign.
It can be expected that Xiaopeng will open the volume, and Huawei, Ideal, and Weilai will soon release corresponding plans. For no other reason, the current intelligent driving is not only a conceptual existence, but also a very important point in the Chinese market to grab sales. As early as the first half of 2023, many 4S stores of new forces have required users to test sales consultants during test drives, allowing users to try the urban NOA function. The results received are amazing, such as the **Xpeng G9 after the big price reduction, and the MAX version of Zhijia once accounted for more than 80%. The same is true of the new M7 of AITO, which has sold out, with the intelligent driving version exceeding 60% and the NCA option rate in urban areas exceeding 70%. Even the ideal version with lidar once broke 10,000 alone.
Of course, most people are still saying that intelligent driving is unreliable, and how to really buy how to IQ tax. But obviously, human nature is irreversible, and no matter how much you say, you can't erase people's yearning for beautiful and unknown things that may happen in the future.
As for how long this yearning can last, unless it is a riotous operation like Weimar and Gaohe, which destroys user trust again and again. Otherwise, as long as you stay up and continue to fulfill the OTA, the results will be good.
As for whether it can become a mass group like new energy replacing fuel vehicles, in fact, look at Huawei now with AITO to ask the world to fly, facts speak louder than words in a short period of time.
The involution, also by this trend, has once again been profoundly cut into two levels at the level of new energy vehicle consumption. The rigid demand layer, the volume is how to reduce the price, how to pile the configuration, generally concentrated within 200,000 yuan. On the upper level, in addition to the cost performance of the volume, the focus has been biased to intelligent driving, focusing on more than 200,000 yuan.
As for Tesla, it continues to be the company that everyone wants to kill it, but its sales are rising, and even if its market value evaporated by 570 billion yuan in January this year, it still makes most companies feel terrified, even if they don't say it.
When talking to ordinary Chinese consumers about keywords such as Tesla, FSD, and autonomous driving, even if there is no word that starts with no or no, the answer will be revealed from body language and micro-expressions. It's not hesitation, it's not confusion, it's mostly resistance.
Moreover, even if they have experienced the high speed of Tesla Autopilot of their friends and family, this resistance will not change or ease anytime soon. In the perception of most Chinese consumers, Tesla is Tesla, and others are others, but this perception is more or less negative.
In the eyes of Chinese automakers and autonomous driving companies, even Huawei, which is far ahead, has almost the same conclusions as consumers.
Tesla is Tesla, others are others. But the emotions behind it are more praise, competition and learning, because regardless of whether FSD Beta has a traffic accident in the world, it is the world's first point-to-point complex road conditions system, its pure visual logic is learning for everyone, and more terrifyingly, its effective data, annotation, learning, execution and other performance. When the computing power is full, when it enters the Chinese market, the impact is so great that it directly forces Huawei to build an autonomous driving zone in Ulanqab, cloud computing big data and other means to prevent it.
FSD will definitely accelerate its entry into China, but when it will enter, Tesla itself will not say, it is useless for anyone to guess, it will be similar to its sudden adjustment. Before writing this article, I called several intelligence-related engineer friends from major mainstream manufacturers, hoping to get cheap words from them, but the answer is the above paragraph, no one knows.
And the speed of its promotion in overseas markets does not have much reference value. In 2019, FSD-related hardware was released, but the actual beta version was launched three years later, in November 2022, and it was only available in North America.
Compared with North America, although European regulators focus on a "lot of things", the implementation speed of FSD is actually not slow. In 2022, Musk set up a flag that Europe must push this year, and after that, there is the familiar bounce. But then in 2023, it will be followed by the recruitment of ADAS test safety officers, and at the same time, HW40 has been approved in Europe, including the user manuals in Europe, China and other places suddenly updated the FSD-related content.
Although there is no evidence chain to piece together when Tesla's FSD, which has already established a team in China, will land, there is a corresponding understanding in the technology circle, that is, in China, the United States will pass relevant regulations, and most of the time will only be half of Europe, or even shorter. As well as the latest news sporadically dug up by foreign media, Europe will land in the first quarter of 2024 at the earliest, and in the third quarter of summer at the latest, while landing in China may be realized in the summer.
It doesn't matter, Musk may skip the ticket again. This is true, but referring to Tesla's announcement at the beginning of 2024 that it would invest $500 million in the Dojo supercomputer, Musk's early response to the decline in profits, financial reports, and stock prices is to buy AI computing chips from Nvidia and AMD. And now Chinese car companies are eager to roll up intelligent driving and grab data. Anyone with a discerning eye can see that the data war has started a long time ago, and even if consumers don't understand, it's better to understand it later.
And grabbing data, in essence, is still highly homogeneous, although now because of intelligent technology and automobiles, but everything still can't jump out of Moore's law and first-mover advantage. Xpeng has the largest continuous investment in intelligent driving among the new forces, the highest proportion of R&D team personnel, Huawei has the longest accumulation of intelligence, self-developed and self-produced sensors, computing platforms, etc., NIO has the latest chip plan, and the ideal is self-developed and carries out L2+ cost reduction applications with Qingzhou Zhihang.
But, all this, pulling, pulling the car, and not getting A is significantly stronger than B, this kind of answer that can make users distinguish, and even keep the same frequency with the logic of online games, a generation of versions, a generation of gods. Tesla uses the hardware developed a few years ago, and it is still not sold. hw3.0 chip has a maximum computing power of 144tops (1 trillion fixed-point calculations per second), in comparison, Weilai's computing power has already broken 1000tops, and the depth of the car rim is now the most advanced, and the next-generation chip Thor delivered in 2025 will break 2000tops in a single chip.
As for the 2000tops computing power to roll up the opponent data, the public level will feel efficient, Tesla hw4The single-chip computing power of 0 chips is expected to be 216 tops, and the maximum total computing power is estimated to be 300-500 ttops. This is the first-mover advantage of self-development and self-production, and vertical integration of outsourcing.
Hardware and software, naturally there are ways to chase and learn, but the data is not good, whoever takes it is who's whom.
The current situation is that the car companies that have their own NOA capabilities in the city have basically caught up with Tesla in terms of software and hardware construction, so it is natural to start rolling up data. Huawei, Xpeng, NIO, and Ideal, this is the time to see that the algorithm framework has basically been completed, and it is now in the regulatory window period and continues to embezzle more data. On the one hand, the data demand of urban NOA according to Xiaopeng Motors before, at least 6 times more than high-speed NOA, if you want to make the car have a selling point, if you want to make the intelligent system take the lead in forming an advantage as much as possible, you have to eat the data as hard as possible, in short, sell more cars, drive more, analyze more, and make more use.
Therefore, from a long-term perspective, even if you sell your car at a loss now, your future earnings are expected to turn red overnight. In fact, this is an important fulcrum of the new forces from the perspective of capital.
The window period lies in the fact that the current regulations on the compliance and data review requirements of new energy vehicle companies in terms of geographic data collection are in a relatively relaxed period around the world. After all, it is impossible for car companies to pay users instead to help run data on the road.
Missing the gap period, forming a structural gap with other companies in key technical elements, it will be difficult to cross in the future. Therefore, taking the familiar internal combustion engine as an example, Chinese car companies have started to learn from joint ventures and reverses, and after 30 years, they can achieve the emergence of the world's leading ones, such as the Geely 2 that has eaten Volvo technology and jointly developed it0TD engine, but can be said to be the world's leading, no.
Well, the answer is quite clear. In 2024, the visible window period of intelligent driving data of car companies will not disappear, after the first line of rolling, the second-line Nezha, Leap and the traditional transformation of Geely, Changan, BYD, etc., this year's goal is not even a small-scale landing, but a large-scale, and even wants to form a standard landing in the level of 150,000 yuan.
Before Tesla's FSD landed, first grab the basic data disk, and after Tesla's FSD landed, use the data more efficiently to improve the system, and then rush to L3 and L4. This is not only the R&D goal of the intelligent driving division of car companies, but also the topic of how to survive for enterprises, and it is also the way for China's new energy vehicles to continue to win in the second half after winning the first half in the world.
Bringing this kind of business back to the car brand side, in fact, for Wei Xiaoli and other brands, everyone's vitality is only once. As for Huawei, because it has more contact points, it has two chances, because even if it encounters obstacles in automobile-related businesses, its competitiveness has declined, and even its profitability is meager. However, Huawei obviously has a backhand.
Because unlike the changeable one-person behavior, giant-level enterprises cannot break out of the cycle of preliminary research, finding the direction of the wind, and planning before making decisions. Looking at Huawei's latest moves, Huawei Technologies has applied for the registration of a number of trademarks such as "Yinwang", "Yinkan", "Yinzhong", "Invisol" and "Innoward", and the international classification includes scientific instruments, means of transportation, etc., and the current trademark status is pending substantive examination.
At the same time, on January 16, Huawei established Shenzhen Yinwang Intelligent Technology Co., Ltd.***, the legal representative of which is Zheng Liying, with a registered capital of 1 billion yuan and a registered address at Huawei's Shenzhen headquarters. The business scope includes intelligent vehicle equipment manufacturing, auto parts research and development, artificial intelligence industry application system integration services, artificial intelligence basic software development, artificial intelligence theory and algorithm software development, communication equipment manufacturing, data processing and storage support services, information consulting services, etc., which are wholly owned by Huawei Technologies.
It is not said that it coincides, but it is highly overlapping. Considering a few time nodes, in 2025, Yu Chengdong's car BU business will face financial assessment within Huawei, according to its previous public interview with **, (car BU) is Huawei's internal latecomer, with an annual loss of 10 billion, and to achieve breakeven, the car-related sales target is 1 million. The newly established company in January has released rumors that it will be loaded with Huawei's current technology and resources, which is currently wholly owned by Huawei, and will become a diversified technology open platform with more partners joining in the future.
It is true that in the business society, there is no shortage of gambling operations that pat the head to make decisions, but when the company comes to a certain size, especially after the global listing, it can carry out such operations, not to mention the basic disappearance, is relatively limited.
The intensity of the crazy involution of the city NOA in front of us is actually worse than the previous fuel transformation of new energy, because the growth of the first business is really fast, even if you don't choose CATL, there are BYD Fodi, China Innovation Airlines, etc. However, the logic of intelligent driving is not like this, and if the data base is robbed, there will be a corresponding competition trend similar to that of smartphones, the stronger the stronger, the weaker the weaker.
For example, Apple's previous delivery results, although the company's shipments and revenue only accounted for 18% and 48% of the global share respectively, but in terms of profits, it can monopolize 85% of the global market.
With He Xiaopeng's XNGP starting global research and development, as all cities in China have been covered by XNGP, Huawei shouted on AITO to ask the world "the whole country can be opened, the more open the better", Li Xiang said that the goal of opening the city was completed ahead of schedule, BYD formed a 4000-person intelligent driving technology team, Chang'an announced the name of Huawei at the beginning of the year, NewCool. Including the negotiation between FAW and DJI, all of this has ignited the fuse of intelligent driving.