In the latest issue of the Ministry of Industry and Information Technology's declaration catalogue, the ideal new flagship model L9 is prominently listed.
According to the official publicity of the ideal, the L9 uses the second-generation platform in the true sense of the ideal, which has made great progress compared with the current ideal ONE.
The biggest change in addition to the second-generation extended range power platform is that it is equipped with the latest NVIDIA automatic driving chip, which has stronger computing power and can achieve higher dimensional automatic driving.
In fact, not only ideal, Niolai's new model also changed the supplier of autonomous driving chips from Mobileye to NVIDIA. Bydir, which has attracted much attention, eventually chose to cooperate with NVIDIA.
In addition, the new forces of Xiaopeng, Weima and Zhiji are also using NVIDIA chips, and Baidu, Didi, and Xiaoma Zhixing are also using NVIDIA's systems.
At a time when car companies are still worried about "losing" their soul in cooperation with Huawei, NVIDIA has quietly stood at the tide of the times. How did it rise? How will its existence affect the development of Chinese car companies? Let the kung fu car take everyone to see it together.
(1) Blizzard and Audi are nobles, and NVIDIA has become an industry giant
The name Nvidia may be unfamiliar to ordinary consumers. But in the digital world, it has long been famous. Because the current CPU has basically reached the limit of energy consumption - performance ratio, whether it is Intel or AMD, relying on the silicon process, it is basically difficult to have earth-shaking changes.
In this case, the image recognition and calculation of the graphics card is particularly important. Not just in the digital realm, but also in the realm of autonomous driving. As the industry leader, NVIDIA has actually locked in the future in advance.
Nvidia was born in 1993, compared to Intel, Microsoft and other giants, it is actually not too old company. In the early days, it was mainly to provide technology services for software giants, such as it developed DirectX drivers for Microsoft, which were used in software for rendering 3D graphics. It then provided graphics cards for Microsoft's first Xbox console and began to have a certain reputation.
The turning point came in 2003, when Nvidia partnered with Blizzard to launch the hugely popular World of Warcraft, whose outstanding visual display effects had a huge impact inside and outside the industry, making its image processing power popular. It was in this year that Nvidia received an invitation from NASA to work with them on a realistic simulation of Mars.
After that, Nvidia's road has become smoother and smoother, including establishing a partnership with Sony, establishing a strategic alliance partnership with IBM, and being adopted by Lenovo. Before entering the automotive field, NVIDIA has long been calling for wind and rain in the Internet field.
In 2015, NVIDIA began to become a graphics chip supplier for Audi. That same year, it unveiled Nvidia Drive chips for smart driver assistance. It is the first to propose the idea of a software-defined intelligent driving car, and the infrastructure is mainly divided into three parts, namely, training the perception, planning, and driving of intelligent driving cars, simulating driving using the DRIVE Constellation simulation platform, and combining developer playback driving data with simulation tests to provide a more comprehensive solution.
The latter things everyone knows, Nvidia gradually occupies an increasingly high position in the field of intelligent driving.
(2) Overturning Mobileye, NVIDIA on what
As mentioned earlier, NVIDIA began to intervene in the field of intelligent driving assistance in 2015, which is not too early. So who was the king of this field before it? The answer is mobileye from Israel.
As early as 2004, Mobileye released the first generation of intelligent driving chip EyeQ1. It solidified its own vision algorithm on this chip, and reached cooperation with first-line auto parts suppliers such as Delphi, Continental, and Denso, and entered the automotive front-loading market early.
More than a decade ago, Mobileye has provided radar image recognition functions for key models such as the BMW 5 Series, so that vehicles can achieve higher dimensional adaptive cruise, which occupied more than 70% of the market at that time, and giants such as GM, BMW, and Volvo were all guests. Including Tesla, before 2016, it also used Mobileye's chip.
It was during this period that Nvidia rose, and a large number of companies that "abandoned" Mobileye were invested in NVIDIA' hands, which is actually positively related to the different concepts of the two.
Mobileye intelligent driving is not only to provide chips so simple, its cooperation model is to package the technology they master, and then make a "black box", car companies only need to buy this "black box" back, and then do a little adaptive development, and adapt to their own vehicles. This is actually a time-saving and labor-saving thing for traditional car companies such as Volkswagen and BMW. But for a new force like Tesla, it seems unacceptable. First, they cannot achieve higher-dimensional intelligent driving, and second, they cannot form a technical gap with traditional manufacturers with deep pockets.
At this time, NVIDIA appeared, which encouraged car companies to develop their own algorithms for intelligent driving software, and mainly provided hardware to match them. Coupled with its inherent advantage in computing power, it is completely a dimensionality reduction blow with Mobileye, and soon occupied the market.
(3) Do not repeat the mistakes of Qualcomm
For domestic manufacturers, choosing NVIDIA seems to be a effortless task. It does not care about the algorithm of car companies, and the price is relatively affinity after the formation of scale, and even BYD, which is particularly inclined to self-research and cost control, has chosen it, which is enough to show its competitiveness.
However, so many car companies "surrender", which also hides very big hidden dangers.
At this year's Electric Vehicle 100 People's Conference Forum, Miao Wei, former chairman of Dongfeng Company and former minister of the Ministry of Industry and Information Technology, specifically talked about the supply of chips. He mentioned that China's automotive industry is still in a state of lack of cores and souls, and it is very irresponsible for domestic car companies to throw key card neck technology such as vehicle specification chips directly to suppliers. He also mentioned a very ugly word, that is, domestic car companies "light calling", few are determined to do these basic research and development.
Not long ago, Yu Chengdong, CEO of Huawei Smart Car BU, also talked about this problem in an interview, he mentioned that the cost of 10 yuan, 20 yuan car chips, now speculated to 2500 yuan a piece, a smart car to use 9 chips, which is absolutely unacceptable to him.
Huawei is really suffering losses in this regard, when your lifeblood is in the hands of foreign companies, they can not only speculate the price to 250 times, but also directly cut off supply at a critical moment, then I am afraid it is too late to cry.
A few years ago, Huawei's Kirin 990 5G mobile phone can achieve a price of 2,000 yuan, after the sanctions, not to mention Huawei, the entire mobile phone industry can no longer buy such a cost-effective product.
But to be honest, it is also unrealistic to expect car companies to stand up and fight with NVIDIA. Nvidia's market capitalization is more than $500 billion, and it is indeed too difficult for young domestic smart driving chip suppliers to compete with it. Especially in this era of rapid technological upgrading, chips that may have just invested a lot of energy to make have immediately fallen behind, and there is no market at all. It is no wonder that enterprises are just "calling", after all, this is a completely unpredictable industry of input and output.
In the view of Kung Fu Automobile, domestic car companies can consider alliances, and everyone sets aside a certain proportion to purchase independent chips, which can be used on low-end cars first to ensure that domestic companies can get certain profits and continue to iterate technology. In this way, you can avoid being stuck in the neck by NVIDIA, and you can also have stronger bargaining power.
(4) Kung Fu shooting
The fact that all the domestic car companies have fallen to NVIDIA is actually a very dangerous thing.
Although Nvidia has not yet shown any fangs, once it is completely enlarged in the field of intelligent driving, it is difficult to guarantee that it will not become the next Qualcomm. In particular, domestic automobiles are rising rapidly, with sales, markets, and funds, and there should be more right to speak.
Huawei's sanctions are always in sight, and we must not let the mistakes that have already occurred in the field of mobile phones and computers reappear in the automotive industry.