Breaking News

Why did chip giant Qualcomm/Intel buy self-driving companies hugely?

On August 5, Qualcomm, one of the global chip giants, officially announced that it had proposed $4.55 billion in cash to acquire Veoneer, a Swedish autonomous driving company, following Intel’s $15.3 billion acquisition of Israel in 2017 After the autonomous driving company Mobileye, the chip giant once again launched an acquisition of the autonomous driving company.

It is worth mentioning that there is also a Tier 1 supplier Magna that hopes to acquire Veoneer, but Qualcomm’s bid of $4.55 billion is much higher than Magna’s $3.8 billion. Although Veoneer rejected Magna’s M&A offer to pay $110 million, it is still in the interests of Veoneer’s shareholders as a whole.

For Qualcomm, it does not hesitate to acquire Veoneer at a higher price, and the CEO of Qualcomm has said that Qualcomm is really interested in Veoneer’s software division Arriver, which has been established since 2020 and is responsible for the research and development of advanced autonomous driving (ADAS). Obviously, Qualcomm hopes to follow Intel and obtain a closed-loop business model of autonomous driving chips + software (autonomous driving algorithms). The war between Qualcomm and Intel has also hit smart cars from 5G.

In fact, both autonomous driving suppliers and vehicle companies, including benchmark companies such as Huawei and Tesla, pay great attention to the business closed loop of chips + software algorithms. On the one hand, they can provide users with the best autonomous driving experience (including OTA). On the other hand, it can greatly increase the value of bicycles, and even achieve the ultimate goal of complete unmanned driving, which may create a RoboTaxi operating fleet.


Why chip giants are buying self-driving companies

In 2017, Intel acquired Mobileye, an Israeli autonomous driving company, at a super high price of US$15.3 billion (equivalent to more than 100 billion yuan at the time). It has raised the imagination of the global autonomous driving company’s valuation ceiling.

For Intel, the overlord in the PC era has to seek new business growth points since the PC business peaked, and the improvement of automotive chips, especially the high computing power requirements of autonomous driving chips, and the PC’s demand for CPUs The demands are relatively consistent, which is why Intel is willing to spend a lot of money on the acquisition of Mobileye, essentially spending money to buy tickets to the future of autonomous driving.

For Qualcomm, the overlord of this mobile phone era, the source of the term Qualcomm tax, has to seek new business growth points after the mobile phone business peaks. It is worth mentioning that Qualcomm’s chips occupy a high market position in the cockpit field, and the world’s first 7nm smart cockpit chip is also equipped with Qualcomm’s third-generation Xiaolong automotive digital cockpit platform.

For Qualcomm, it is not enough to win the chip market in the entertainment field. It also needs to enter the market in the driving field. The chip computing power is the same, but the software algorithm is completely different. In particular, Apple CEO Cook believes that the global The artificial algorithm of self-driving, the mother of artificial intelligence, so Qualcomm is willing to wave the check to buy the self-driving company Veoneer, which is not much money for chip manufacturers who have been in a full-scale boom in the chip industry recently.

Whether it is Intel or Qualcomm, both hope to open up new markets when their main businesses have peaked, and they all take a fancy to the opportunity of autonomous driving. This is a huge market opportunity brought about by intelligent vehicles, and it is also a global Tech giants, including Huawei, Apple, Google, and Microsoft are targeting new markets.

Moreover, these chip giants all believe that it is not enough to have chips alone, but also to add autonomous driving algorithms to obtain greater bicycle value. Intel’s acquisition of Mobileye, and Qualcomm’s attempt to acquire Veoneer, is the logic. This logic also has a very strong practical support, which is the case of Tesla.

In fact, the domestic chip company Horizon is also trying to commercialize the chip + software algorithm, and even at the press conference of Journey 5 in July, it will build the Together OS in-vehicle operating system. This is very ambitious. It depends on whether we can leverage the basic hardware capabilities of large computing power chips to leverage the basic software capabilities of in-vehicle operating systems.

Why did Intel choose to acquire Mobileye and Qualcomm choose to acquire Veoneer? Strictly speaking, these two companies are currently in the field of assisted driving, but this is also the most profitable field in the field of autonomous driving at present, and the commercialization of L4-level autonomous driving , it still takes time. More importantly, Mobileye and Veoneer are already suppliers to car companies. Intel and Qualcomm only need to develop chips and then improve algorithms to supply car companies, thereby generating cash flow.


Commercial closed loop of chip hardware + software algorithm

In 2020, Tesla’s financial report shows that FSD contributed more than $1 billion in revenue, which is still achieved on the premise that Tesla’s US car owners’ opening rate is 20%, and the Chinese car owners’ opening rate is less than 5%. The performance of Tesla’s FSD has made the global autopilot companies ecstatic, and they are full of energy to do it.

In Tesla’s FSD business model, a very important approach is that hardware is standard and software is paid. The so-called standard hardware means that Tesla’s vehicles are equipped with hardware devices such as FSD chips and sensors when they leave the factory. When users need to activate the FSD function permission, they only need to activate the software. This is the Internet business model that the automobile industry has dreamed of for a century, and the price of FSD at $10,000 is already very high.

For the sake of revenue, Tesla also sets up a mode of different functions and different payment standards. Tesla AutoPilot is standard, that is, it does not require extra money. 40,000 RMB, Tesla has also launched a monthly subscription payment model in the United States. In the near future, Tesla will launch a similar payment model in China after pushing the FSD V9.0 version, and may even follow Pay for different functions and different durations of use.

The basis of all this is because Tesla has built a system of autopilot chips + autopilot algorithms. Because of this, Tesla can continuously provide more autopilot capabilities through OTA, and can develop more autopilot capabilities. charging model. This is also what Intel hopes to do after its acquisition of Mobileye and Qualcomm’s acquisition of Veoneer.

Why did chip giant Qualcomm/Intel buy self-driving companies hugely?

Huawei’s autonomous driving capability also hopes to package chips + algorithms. Previously, Huawei MDC would not sell chips alone. However, recently, Huawei MDC has begun to cooperate with autonomous driving companies to provide auto companies with autonomous driving capabilities, such as , Huawei MDC has cooperated with MINIEYE to provide such a solution for passenger car companies.

However, there is no doubt that this is a huge challenge for Huawei, Intel, and Qualcomm, because for car companies, if they want to achieve the level of Tesla, they need the cooperation of the whole product life cycle. , This requires car companies and suppliers to work together to do a good job of data ownership and interest division. Compared with Tesla’s model of doing everything by itself, the supplier’s model is obviously a big challenge. But it is also a huge opportunity.

Mobileye sold $15.3 billion because Intel needed to supplement its autonomous driving capabilities, and Mobileye had a large number of pre-installation orders from international car companies and certain chip capabilities for autonomous driving. Veoneer sold $4.55 billion, It is also because Qualcomm needs, and also holds fixed-point orders from car companies, which may also be a reference for companies of the same type in China now. For companies aiming for L4 autonomous driving, the same can be used as a reference.