The special nature of the automotive industry means that it is almost impossible for a single company to dominate the supply chain. The same is true for the chip track.
Taking intelligent driving as an example, the current supply chain form is also in a critical cycle of change. Players from all walks of life are also trying to diversify their cooperation models with car companies.
According to the latest monitoring data of Gaogong Intelligent Vehicles, NVIDIA is currently recruiting a large number of software engineers for autonomous driving solutions in the Chinese market, with offices in Beijing and Shanghai, with the goal of deeply participating in validating, adapting and improving the performance of the software stack in China.
Recruitment directions include high-definition mapping and positioning, behavioral decision-making, planning and control, system solutions and automated parking system solutions, and also launched a "full stock ownership" plan.
This means that the leading chip manufacturers are strengthening the "software and hardware synergy" strategy.
In the era of traditional automobiles, chips are attached to Tier1 service car companies; Different Tier1s, due to the tendency of solution selection and the degree of close cooperation, lead to different chips have different "owners".
For example, in the early era of assisted driving, Bosch and Veoneer chose Xilinx, Continental chose TI, Valeo, Aptiv, and ZF chose Mobileye.
Entering the intelligent cycle, late entrants such as NVIDIA and Qualcomm have seized the "computing power" and sprung up to become the new backers of the transformation and upgrading of leading car companies and Tier1.
However, the startups that followed the entry are not weak. Even Mobileye, which enjoyed market dividends exclusively in the first wave of assisted driving, once had a market share of up to 70%, may become a luxury.
Cost and cost performance have also become the focus of Mobileye's attention.
When talking about the current fierce competition in the market, Mobileye's senior executives believe that "the performance and cost of the system are the key to determining who will win in the end." ”
The company also pointed out that as some joint venture brands in the Chinese market and China's local new energy vehicle companies continue to seek breakthroughs in high-end intelligent driving technology, "we are facing increasing competitive pressure." ”
In the high-end intelligent driving track, limited by the price range of the model (cost consideration), NVIDIA's monopoly is also becoming a thing of the past. Since last year, cost-effectiveness has become the mantra of car companies and Tier1, which has also given birth to different solutions for models at different price points.
The current market form is that some leading car companies and new forces Tier1 are seeking to develop their own chips, and the success of Tesla's model is the main factor. For example, Ideal, Momenta.
Most market players have begun to reserve solutions with different costs, such as Nvidia, Qualcomm, Mobileye, TI, Horizon, Black Sesame Intelligence, Aixin Yuanzhi and other diversified choices, which have also intensified market competition.
However, due to the large R&D investment and long cycle of the chip industry, the traditional industry logic is to rely on scale to achieve sustainable profitability and R&D investment in next-generation technologies.
One of the paths is to collaborate with software and hardware, from hardware to software, to obtain a higher unit price. After all, for the automotive industry, the biggest difference from consumer electronics is that the ceiling effect is obvious.
For example, for Nvidia and Qualcomm, the automotive chip (hardware) business has gained momentum in the past few years, but the contribution to the overall performance of the two companies has not yet reached a critical point.
According to the data, in 2021, NVIDIA's automotive business revenue will be 566 million US dollars, accounting for 2% of the company's total revenue, and in 2022, the revenue will be 903 million US dollars, accounting for about 3%. Although Nvidia has more than $10 billion in orders in hand, it is still unknown whether it can be delivered.
For example, in many brand models, the strategy of high and low configuration is also very obvious. Whether it is ideal or BYD, it has its own alternative to NVIDIA's configuration strategy.
For latecomers, how to help car companies reduce software development costs is also one of the key factors to increase the "winning rate". After all, the proportion of self-research by different car companies varies greatly.
In April 2022, Qualcomm partnered with private equity SSW Partners to complete the acquisition of Veoneer for $4.6 billion. Next, SSW Partners will spin off its Arriver (software asset) subsidiary to Qualcomm.
At the time, Qualcomm CFO Akash Palkhiwala said that Arriver's ADAS software division would be an important part of Qualcomm's rapidly growing automotive business. At the level of Qualcomm's overall strategy, it is also a milestone in the diversification of its business model.
A Qualcomm spokesperson made it clear that "for some automakers, if the software development capabilities are weak and the development costs of third-party traditional Tier1 suppliers are too high, we can directly provide a complete ADAS turnkey solution." ”
And for Nvidia, the same is true.
Nvidia made an analogy, "The reason why our data center business is successful is that in addition to the hardware itself, we also provide development platforms, software, SDKs and other resources to help customers quickly land." ”
Previously, NVIDIA introduced the NVIDIA DRIVE™ open-source software stack, which enables developers to efficiently build and deploy a wide range of applications, including perception, localization and mapping, planning and control, driver monitoring, and natural language processing.
In addition, NVIDIA offers a modular configuration of software, including NvMedia for sensor input processing, NVIDIA CUDA® libraries for efficient parallel computing, NVIDIA TensorRT™ for real-time AI inference, and other developer tools and modules with access to the hardware engine.
To this end, NVIDIA has acquired DeepMap, a high-definition map start-up, to strengthen the market competitiveness of Nvidia Drive by integrating the latter's technical solutions. "They will expand our global mapping business and expand the solution capabilities of our fully autonomous technology stack."
Three years ago, Nvidia and Mercedes-Benz announced that they would collaborate on the development of fully autonomous vehicles over the next decade, drawing on their respective experiences in high-performance computing and high-end vehicle manufacturing to create a new software-defined vehicle.
This cooperation is regarded by NVIDIA as the "largest single business model transformation" in the company's development history. The two parties will be able to share the revenue from future user purchases of features and subscription services. Behind it, NVIDIA provides full-stack software and hardware solutions, and it is jointly developed with Mercedes-Benz.
To this end, Huang also emphasized that "in the next few years, it will no longer be impossible to sell new cars at cost price, because profits will come mainly from software." "Public data shows that Nvidia will earn $8 for every $1 sold to hardware, software and other related system providers.
Previously, Huang also proposed the concept of "software 3.0" (which relies on data, algorithms and compute engines), which, in conjunction with the company's strategic transformation, said, "Nvidia wants to sell more software that only runs on its GPUs." ”
Another milestone is that this year, Wu Xinzhou, vice president of autonomous driving at Xpeng Motors, announced that he would join NVIDIA, where he completed the mass production of a full-stack high-end intelligent driving solution at Xpeng Motors.
It is reported that at present, the strategic cooperation project between Nvidia and Mercedes-Benz has been postponed due to the lack of full-stack software capabilities of the former, coupled with unexpected factors such as the epidemic in recent years. In addition, Mercedes-Benz is even seeking help from other suppliers for its projects in China.
For example, Mercedes-Benz invested in Momenta, the company's high-end intelligent driving is also based on the NVIDIA platform, and has begun to achieve mass production and delivery of a complete set of software systems in the Chinese market.
This means that NVIDIA needs to come up with a full-stack solution that can be implemented as soon as possible. As the current market share leader in the high-end intelligent driving track in the Chinese market, NVIDIA's "sense of crisis" is also highlighted.
This week, ECARX announced that the Tianqiong Pro integrated driving platform equipped with the Lynk & Co 08 has been mass-produced and delivered. The platform is available in two configurations, including "Sky Dome Pro" and "Sky Dome", both of which support NOA Pilot Assist.
The Sky Dome platform is equipped with the Black Sesame Huashan No. 2 A1000 chip, which can provide a maximum of 58TOPS (INT8) computing power, and adopts 6 cameras + 5 millimeter-wave radar solutions to support high-speed piloting.
The Tianqiong Pro platform is equipped with two A1000s, with a comprehensive computing power of 116TOPS, supporting BEV perception algorithms and 5R10V sensor solutions (optional lidar), which can realize high-end functions such as urban NOA pilot driving assistance and HPA memory parking.
Previously, public information showed that the BOM cost of the integrated domain controller of Black Sesame Intelligence, which supports the 10V (camera) NOA function, is controlled within 3,000 yuan, and at the same time, it is still continuing to promote cost optimization and has the opportunity to further explore.
According to the latest data released by the Gaogong Intelligent Vehicle Research Institute, from January to September 2023, 377,300 new cars were delivered to the Chinese market (excluding import and export) passenger car front-loading standard (software and hardware) NOA, an increase of 151.20% over the same period last year.
However, in terms of front-loading rate, it is only less than 3% (2.55%). This means that before hitting the 10% inflection point, there are still various uncertainties in the market, such as the allocation of solutions.
The picture and text are from the Internet, not for any commercial use, if there is any infringement, please contact to delete. The content is for reading purposes only and does not constitute investment advice, so please treat it with caution. Investors act accordingly at their own risk.