Commenting on media reports that NIO is planning to develop a self-driving computing chip, experts in the car industry said it was difficult to predict the future of the move.
Although NIO's move is conducive to building core competitiveness, it also faces greater technical and financial risks, Caijing.com.cn quoted industry sources as saying.
"At present, only Tesla has built a systematic technological advantage from the underlying chip to the application level, but it has also gone through a longer iteration to finally become competitive," the report said.
Automotive analyst Ren Wan pay said, including NIO, the reason for the car companies to develop their own chips, is to take the initiative and advantage as far as possible, in order to build the enterprise's "moat", but "NIO self-developed chips, good icing on the cake, do not do well may become a burden. "
CMIB International Research Department Bai Yiyang said the auto driving chip belongs to the key technology of car enterprises, directly determine the performance of automatic driving, involving the security of the supply chain of car enterprises, currently monopolized mainly by foreign companies such as Mobileye, NVIDIA.
Domestic Horizon.AI and other companies, although starting late, but the momentum is strong, and is expected to replace some imported products.
"It's too early to judge the prospects of NIO's self-developed chip, which, if successful, will be a major milestone in China's self-driving hardware," Bai Yiyang said.
The model of car companies to develop their own chips was pioneered by Tesla.
In April 2019, Tesla released the FSD fully automated driving computing platform, replacing the previously used NVIDIA chip, equipped to enter the Autopilot 3.0 system.
According to Tesla, the FSD chip has 144 TOPS of computing power, nearly five times that of the NVIDIA Xavier, enough to support the computing power required for full autopilot.
Tesla has also adopted Mobileye's EyeQ3 and NVIDIA Drive PX2 before landing on its own chip.
With the introduction of FSD, Tesla has strengthened its position in the industry from the perspective of technology and cost control, and become a model for new carmakers, including NIO, to copy.
Today, many automakers, including NIO, Xiaopeng and Ideal, have set up self-development teams for ADAS technology.
For NIO self-research, automotive analyst Zhang Xiang said bullish. He believes that the car enterprises "self-developed chip" is usually just a design chip, which constitutes the basic unit of the chip is mostly purchased, the chip production process rely on commissioned processing.
"Tesla from the cooperation with Mobileeye, NVIDIA, etc., to the launch of FSD, it is two or three years, the final production is entrusted to Samsung, NIO research and development is completed in China, the cost is likely to be lower than Tesla." Zhang Xiang said.
In fact, NIO is very early in the layout of the smart network connectivity, including automated driving technology.
Back in 2015, NIO hired Padmasree Warrior, a famous Silicon Valley entrepreneur, as CEO of NIO North America to take charge of NIO's global software development and overall management of the North American company.
Warrior left the company in 2018, and by the end of 2019, the total NIO North America staff shrank from 640 to 141, and the CEO of the North American headquarters left as well.
After layoffs and the departure of several directors, the trend of "localization" of NIO's own R&D teams has become more and more apparent.
Currently, NIO's Chinese R&D team mainly uses Mobileye's chips. At the Chengdu Auto Show, William Li Bin released a signal to the media to develop its own chip.
In August this year, Ren Shaoqing, former Momenta R&D director, joined NIO as assistant vice president, reporting directly to William Li Bin. At the same time, NIO announced plans to target financing of more than $1.7 billion in stock issuance, the proceeds will be used for research and development of automated driving technology.
Bai Yiyang said that the research and development of automated driving chip in at least 3-5 years, the investment should be in the billions to 10 billion yuan magnitude, at present, do not rule out the NIO to use their own research chip plan to start financing the case.