Where and who is involved in a major partnership agreement can often reveal a lot.
On the evening of November 25, Huawei and Changan Automobile signed the Memorandum of Investment Cooperation. Although it is an intentional agreement, important operators from both business lines, including the first, second and third executives, have come, and the location is chosen in Shenzhen. This reveals that Huawei is both a landlord and an important leading party.
The content of the agreement, a day or two before the official public, there have been many reports. In short - when Huawei sets up a new company, Changan will hold no more than 40 per cent of the shares. The previously rumored "Huawei car BU business divestiture" has not been announced, but from the announcement, Huawei promises not to start all over again ", no longer participate in the "target company" business with peer competition, no accident, Huawei car BU will be separated independently and assets injected into the new company.
It doesn't sound very different from the agreements the two sides have signed for the past five years. But as long as you consider that when Avita (Changan, Huawei and Ningde Times built by the three parties) was established, Huawei insisted on not investing and not occupying shares, and did a typical HI model, it is not difficult to understand. As for the car BU stripping from Huawei's system, it is more inferred that this is the business architecture design that wins the trust of the Oems. After all, in the past HI model, because Huawei is only a core supplier, the market feedback is more from its partner brands, such as the previous Celis and BAIC, there are too many variables and uncertainties.
The establishment of Huawei Car BU, like a strong cheetah, rapidly tore open the customary boundaries of the automotive field. But inside Huawei, this is the only business line in the business group that loses money. Long-term doctrine to adhere to, but no one wants to make a loss of conspicuous package, so, want to profit how to do? It is nothing more than expanding the cooperation base and making shipments up. There are dual goals in this, economies of scale, more importantly, equipped with Huawei family bucket products on the road to run much more, Huawei ADS cloud computing power learning miles (non-virtual training), can be greatly increased, which accelerates the speed of model training and transfer. Huawei Cloud's hardware is not a problem, and its monthly computing power increment (about 0.5E FLOPS) is almost equal to the level of retention power of rivals.
The new company is like a "super platform" with three goals: one is joint procurement, the second is joint product architecture design, and the third is to act as a base for computing power for partners. Huawei is already a major producer of lidar, millimeter-wave radar, computing chips, and front-end chips. Software from OS to dual intelligence application software, and then to a variety of basic libraries, tool chains, everything. Simply put, it is to become the largest supplier of double intelligence by virtue of Hongmeng Car machine and ADS intelligent driving.
Now the vehicle E-E architecture and local computing power, as well as sensor hardware, are very closely integrated with software. The functional portfolio solution that traditional Tier1 relies on is already under threat. Oems are increasingly playing the role of Tier1 themselves, responsible for the integration of the whole vehicle into several domain schemes, parts are directly purchased from Tier2, and Tier1 is "short-circuited" in some scenarios.
The explosive point of Huawei this time, not only provides the whole vehicle area program, computing power and software program, but also hardware procurement, and put their own intelligent hardware production capacity on the table, shareholders can take priority to guarantee the supply of goods, even if the price rises, with equity can also hedge.
And the base of this piece of computing power, Huawei is still crazy to pull customers. Internal executives have said that Huawei wants to be the "second choice" of the world's computing power and serve as the base of China's computing power. This is equivalent to doing the work of Qualcomm, Nvidia, Google, and ChatGPT.
Relying on Huawei cloud computing power, the model can be easily trained and transferred (to the car end). You don't have to re-build computing power (think about how expensive the H100 is to buy one), and you can share training data with partners. The ability of the "super platform" is realized as a rapid iteration of intelligent driving capabilities, which may be faster than the current mobile phone update. And based on the complexity of hardware and tasks, OTA updates bring new experiences that are comprehensive beyond mobile phones.
In fact, Huawei in the automotive industry has been Tier1, Tier2 identity, intelligent car selection business can be counted as Tier0.5 (participate in product definition, styling design, marketing, user experience and other links). To some extent, it is not an exaggeration to call Huawei a "super supplier to the automotive industry". There has long been a view that Huawei intends to do "another Bosch", but by careful comparison, Bosch's tentacles extend to Tier2 at most, and it has never intervened in vehicle design, let alone customer marketing. In the new energy era, there is no high-level assisted driving program, and there is no high-speed NOA.
There is no doubt that Huawei and Changan have created a new model. Just as in the era of fuel vehicles, Tier1, such as Bosch, and mainstream automobile manufacturers in the world to establish a stable cooperative relationship, and even multinational automobile companies will be important first and second tier suppliers cross-shareholding, and under the stable cooperative relationship, everyone shares the principle of profit distribution under the stable relationship.
The upcoming joint venture not only penetrates all of Huawei's automotive business models, making the boundary between HI and smart car models blurred, but also, to some extent, making the vow of "Huawei does not build cars" more convincing. More importantly, this platform is likely to incidentally derive a joint procurement model, intelligent hardware procurement chain, first placed on the platform, several major shareholders will also put the corresponding assets up, can achieve the intensive effect on the cost. Previously, there has never been a similar intelligent hardware collection platform, and new rules in the new energy era are likely to emerge.
The only variable is the number of Oems on board. At present, in addition to the exposure of Selis and Chery, there may also be BAIC and Jianghuai. And HI business is currently only Avita (before the polar fox also counts), HI model before also need to work hard, now as long as on the super platform, enjoy the benefits of the big league, the number of partners (at this stage are Chinese brands) is not worrying.
There is no player on the table, then the speed of the distance is likely to be measured in days...