With the development of new energy vehicles in recent years, automobile "four modernizations (intelligence, networking, electrification, sharing)" has become a competitive place for major automobile enterprises and parts manufacturers, which do not hesitate to invest a lot of money, among which there are many start-up companies. Huawei's entry will undoubtedly make the already competitive industry, the competition more fierce.
The field of autopilot alone has brought together dozens of players, including international giants. If you look at the pattern of autopilot, it is possible that all the companies will not be able to avoid competition with Huawei.
In the era of fuel vehicles, domestic cars "heart disease" is difficult to get rid of, engine control technology depends on foreign suppliers, so that it can not calm the pain. Up to now, the engine control system still depends on the supply of the joint venture factory, and the new technology depends on the foreign side.In the era of intelligent network automobile, chip is the highland that can not be lost in China's automobile industry.
Huawei Hayes series chips in consumer electronics, although can not compete with Qualcomm, Apple A series, but also a small achievement. With years of investment in the IC industry, Huawei Sheng Teng system AI chip, from the public information, up to 352TOPS super computing power has been in a leading position in the industry. Huawei has built a MDC (MobileData Center) platform on this basis, trying to provide auto-driving full-stack full-scene services for automakers.
Huawei Sheng Teng 910 chip, single chip computing density exceeds Nvidia.
In the high-level autopilot, the discrete electronic control unit (ECU) can no longer be used to match each function, but uses the partial centralized domain controller (DCU).
DCU is responsible for the calculation of multi-sensor fusion, location, path planning, decision control, wireless communication and high-speed communication. In the centralized and hybrid architecture, DCU also has to undertake all or part of the sensor data processing. Because of the need to complete a large number of operations, DCU generally has to match a core computing power processor, that is, the chip, to provide autopilot with different levels of computing power support. The higher the computing power, the more functions it supports, so everyone pursues high computing power. There are many programs in the industry, such as NVIDIA, Infineon, Renesar, TI, NXP, Mobileye and so on.
Although the research and development of the chip requires a lot of money, but the chip as the core component of autopilot, its high price and considerable profit, still attract many giants to invest in it. Among the many players are consumer electronics semiconductor giants, traditional automotive semiconductor suppliers, and new players. I chose a few representative enterprises to do a simple comb.
1. NVIDIA: Nvidia has been committed to the development of AI chips in recent years, with a full range of solutions in the field of autopilot:
1) DRIVE PX2 open artificial intelligence vehicle computing platform;
2) NVIDIA DriveWorks software development kit;
3) Digital cockpit, high-precision map, advanced auxiliary driving solution.
2. Intel 's former PC era CPU King Intel missed the smartphone era for fear of missing out on the autopilot era again.After the launch of the IntelGo platform, Mebileye, the dominant autopilot open platform in ADAS vision, also went on sale in December 2018.
3. Qualcomm: Qualcomm's layout in the automotive industry is not limited to the communications areas it is good at, but also in the ADAS and cockpit entertainment domain 820A. At CES 2019, DriveAutomotive, an extensible autopilot platform based on Snapdragon chips, was unveiled for automakers to add modules to their needs.
However, only the layout of the product side does not seem to show Qualcomm's confidence in the entry industry, and NXP's acquisition is also twists and turns, but ultimately fruitless.
4. NXP: as a traditional supplier of automotive semiconductors, NXP aims at a number of control domains, including: connectivity, body comfort, autopilot, information entertainment, powertrain.Among them, the BlueBox computing platform is built in the autopilot domain, and its BlueBox2.0 supports L4~L5 autopilot requirements.
5. Renesas: Riza Electronics's R-Car V3M high performance image recognition system (SoC), can greatly optimize intelligent cameras, panoramic look systems, lidar and other applications. At the same time, Riza also launched the Renesas Autonomy open platform ADAS/ autopilot field.
6. With its own advantages, TI: Texas Instruments cut into the field of low-level ADAS, its vision, ultrasonic, millimeter wave radar and other sensor chip layout, as well as the extension of Jacinto series SoC from information and entertainment systems to ADAS, to launch TDA series SoC, Occupy the market quickly and at low cost. In the field of subsequent autopilot, TI may continue to work.
7. Tesla:Tesla is not willing to just build a car. It recently announced that its autopilot chip FSD, will be carried on the Tesla AutopilotHardware platform in the future. From the point of view of hardware integration, Tesla completed the integration from chip to board level, system integration, and then to the whole vehicle, and opened up the layout of the autopilot industry chain.
8. Horizon: the horizon, which has been established for less than four years, is popular with capital markets and valuations have soared. In 2018, relying on its software and hardware combined with AI processor technology, the company successively released the Matrix autopilot computing platform and the Horizon XForce edge AI computing platform.
9. Cambrian: Cambrian deep learning chip Cambricon-1M, supports personalized deep learning and can be used for real-time multi-channel video processing, including autopilot.
10. West well Technology: West well Technology started from chip development, and in October 2017, United Zhenhua heavy Industry announced the application of its self-driving brand Qomolo, in unmanned heavy truck port ring.