The semiconductor industry, which has long been eyeing the automotive market, has not stopped for a moment. Intel has once again taken steps to acquire computer vision chip developer Mobileye for $15 billion, and NVIDIA will not be outdone, and will release it with Bosch. Create an artificial intelligence system.

Deconstruction of the traditional vehicle industry chain, the proportion of semiconductor plant investment increased

The pace of semiconductor manufacturers is becoming more and more active, highlighting the development of self-driving and car networking, and is deconstructing the industrial chain of the traditional vehicle industry step by step.

In 2020, many manufacturers have set up a key year for realizing the era of self-driving. To welcome the era of driving, the construction of ADAS (Advanced Driver Assistance System) and the Internet of Vehicles constitutes a necessary condition. In the past, the mechanized components and system design of vehicles have been unable to meet the trend of the industry, and it is necessary to improve the control accuracy and gradually make the electronic components of the vehicle electronic. From the vehicle's navigation and entertainment, such as the vehicle's TelemaTIcs products, which are not directly related to safety, to the reversing radar, image recognition system and ADAS related systems, the safety sensing products are all proof of the vehicle's gradual electronicization. . Therefore, from the semiconductor to the electronic component manufacturers, the investment in the development of automotive electronics and the proportion of business is gradually increasing.

Observing the development of semiconductor manufacturers in the automotive market, TI, Renesas, NXP and other manufacturers have already worked closely with the car manufacturers to play a key role in the automotive electronics supplier. In the past, semiconductor chips companies Intel, Qualcomm and NVIDIA, which focused on PCs and mobile devices, did not intend to hand over this market pie. In order to seize the main battlefield of the central processor of the ADAS system, the three companies have been working with the car manufacturers for nearly a year. Alliances, or even large-scale mergers and acquisitions, are aimed at perfecting the technology used in the future of technology, and the ability to create "Computer Vision" has become a consistent goal of these latecomers.

Why is computer vision highly valued?

It is the key to computer vision technology that gives computers the ability to visually interpret and respond to pictures. In terms of the vehicle market alone, with the increasing resolution of the lens and the increasing amount of image data in the ADAS, semiconductor manufacturers must also provide processors with more powerful computing power.

Broadly speaking, due to the processing of visual signals, whether it is the X86 architecture or the ARM architecture Cortex-A series processor, this is not a good project. If you want to deal directly, it will not be efficient. Better, waste more CPU core resources and other situations, thus dragging down the performance of the overall processor.

Therefore, whether the X86 architecture dominates Intel or ARM, the technology of other companies is acquired to complement the original architecture in dealing with computer vision.

Intel's acquisition, alliance, and technology development are aimed at computer vision technology

Let's take a look at the three companies that Intel acquired in the past year. In May last year, Intel acquired Itseez, whose technical expertise is computer vision. The algorithm developed by Itseez allows the vehicle to have the visual ability to distinguish obstacles from collisions. Its driver assistance system can detect vehicles that appear from the street or pedestrians who are crossing the street.

In September last year, Intel acquired Movidius. The company's unique technology lies in mobile image processing technology. Movidius's VPU chip, before being acquired by Intel, Movidius is a competitor of CPU and GPU vendors. Its VPU products are optimized for computer vision and provide powerful visual computing capabilities.

Intel recently re-launched the acquisition of Mobileye, which includes the in-depth technical advantages of machine vision, deep learning, data analysis and high-precision graphics on ADAS products, and then highlights Intel's determination to lay out computer vision technology.

Observing Intel's technical capabilities, before the acquisition of Altera, the inherent architecture of the CPU, it is not easy to cut into the application market outside the PC and server, until there is a very versatile product such as FPGA, open up more market opportunities, Nowadays, Intel’s mergers and acquisitions in recent years are highly related to computer vision. From the chip to the algorithm, Intel’s next step is how to properly integrate its product lines. Cut into the self-driving market faster. This highly integrated advantage is rare among many automotive semiconductor companies.

Qualcomm and NVIDIA's layout strategy in computer vision

Qualcomm demonstrated his determination to enter the automotive market through the acquisition of NXP. Qualcomm, which originally had CPU, GPU and DSP (digital signal processor) independent development capabilities, of course, intends to use its DSP technology to create enhanced computer vision processing, its new generation chip Snapdragon 835 equipped with Hexagon 68DSP, through the internal heterogeneity of the processor The architecture's co-processor capabilities will make computer vision work more efficient and accurate.

NVIDIA is leveraging its own GPU capabilities to continuously improve GPU computing power to meet computer vision or other applications that require high performance computing. Recently, NVIDIA also introduced a new processor called Xavier. The processor uses a 16nm FinFET process, the CPU uses an eight-core architecture that customizes the ARM v8 instruction set, and the GPU is a new generation of Volta with 512 GPUs and a total computing performance of 20 TOPS (trillion operaTIons per second).

In summary, one of the keys to achieving a self-driving vision is to have high-resolution image signals with powerful computer vision computing power. From a chip perspective, to strengthen the lack of computer vision, in order to get the right to speak in the self-driving market, whether it is Intel, NVIDIA, or Qualcomm, have taken different ways to meet this demand, and each with Different car factory partners have launched a self-driving cooperation plan, and in the future, who will be able to take the lead in the market.

500 Puffs

Disposable electronic cigarette

Our products include disposable e-cigarettes, rechargeable e-cigarettes, and various of flavors of cigarette cartridges.We have our own professional team and competitive quotations for any OEM or ODM works.

We supply OEM disposable vape pen,OEM disposable electronic cigarette,ODM disposable vape pen,ODM disposable electronic cigarette,OEM/ODM vape pen e-cigarette,OEM/ODM atomizer device.

12 Flavors : Guava .Mango Ice .Passion Fruit .Mixed Fruit .Blueberry .Litchi Ice .Mango Ice .Lemon .Grapes.Watermelon .Banana Ice .Coconut

Disposable E-Cigarette,Electronic Cigarette Pod Pro,Atomizer Device Vapepod Pro,Atomizer Device Zgar Pod Pro

Shenzhen WeiKa Technology Co.,Ltd. , https://www.szvape-pods.com