Black Sesame Intelligent became the first self-driving chip company invested by Bosch in China. This is also the second time that this young company has received a very high endorsement after being invested by Xiaomi last September.
This is not the first domestic car chip company to enter the public's field of vision. In fact, although NVIDIA, Qualcomm and Intel have made great efforts in the field of automobile master control chips in recent years, products such as Horizon Journey Series and Huawei Upgrade Chip are also emerging.
For example, UNI-T, the main model of Changan Automobile in 2020, chose Horizon Journey 2 automobile smart chip.
Regarding Bosch's strategic investment, Yang Yuxin of Black Sesame Intelligent CMO interprets it this way:
"With the development of the entire automobile industry, even these global iron fists need to adapt to the needs of the times and the local market. Using the local supply chain is actually one of their directions. "
In an understated tone, it seems to reveal the confidence that local enterprises can also lead the development of car gauge chips.
Looking at the mass production and loading of main control chips in China market, the leading forces of mass production are still foreign companies such as NVIDIA, Qualcomm and Intel.
Mass production and loading of main control chips in China market
In April last year, Black Sesame released Huashan II A 1000 Pro autopilot chip. The computing power of INT4 of a single chip can reach 196TOPS, and the model is expected to be mass-produced by the end of 2022 at the earliest.
If the big computing chip of Black Sesame is successfully loaded in 2022, there will be one more independent player. Horizon Travel 5 Series High Computing Chip will be mass-produced in 2023. It seems that local companies have indeed broken the competitive barriers of automotive chips and may usher in a golden period of development.
Affected by the lack of 202 1 core, it has become industry knowledge to master the core technology in the supply chain independently. Under the trend of software-defined automobile, the main control chip and computing platform have become the core of automobile intelligent development.
If the local core enterprises rise successfully, it will undoubtedly add a powerful shot in the arm to the dream of overtaking in corners of China automobile industry.
How will the rise of local companies evolve? Can they destroy the market?
1In mid-October, Dr. Deng Kun, Vice President of Intelligent Application Engineering in Beijing, Yang Yuxin and Black Sesame, respectively expressed their views on the development of the industry.
What is the level of local on-board chips?
Constantly improving the computing power of chips is very common in the competition in the era of smart phones, and now this scene is also staged in the field of smart cars. In the past two years, some car companies have begun to take computing power as a propaganda point:
Weilai ET7, equipped with NVIDIA DRIVE Orin chip, with computing power of1016 tops;
Smart cars and SAIC R cars also use NVIDIA chips with a computing power of 500- 1000+TOPS. ...
To compete for the competitiveness of local core enterprises, we must first ask what level their computing power is in the industry.
Let's compare Huashan II A 1000Pro of Black Sesame with two high-performance vehicle-grade chips that have been listed:
Tesla FSD adopts 14nm process, and its single chip computing power is 72 TOPS.
NVIDIA Orin adopts 7nm process, and its single-chip computing power is 254 TOPS.
Black Sesame Intelligent A 1000Pro adopts 16nm technology, and the single chip computing power of INT4 is as high as 196TOPS, and that of INT8 is 106 TOPS.
The computing power of Tesla FSD and NVIDIA Orin are calculated according to the INT 8 standard, so it can be roughly understood that the position of Black Sesame Intelligent A 1000 Pro is between Tesla FSD and NVIDIA Orin.
However, according to Yu Kai, the founder of Horizon, the pursuit of pure computing power breakthrough is unsustainable, and it will meet the ceiling. Although computing power is important, it does not represent the real performance of automotive smart chips. Therefore, the true level of A 1000Pro will not be known until mass production and loading.
Heizhima Intelligent CMO Yang Yuxin
Although computing power does not determine the final performance, in Yang Yuxin's view, local companies have their unique competitive advantages.
Compared with NVIDIA, Black Sesame Intelligence is closer to customers, and the timeliness of Black Sesame Intelligence will be higher in receiving customer demand and two-way feedback during the development process.
Compared with Mobileye, the market leader of ADAS, Mobileye has always adopted the integrated solution of sensor+chip+algorithm binding, which can rapidly increase the market share at L2 level of autonomous driving. However, in the long run, the "black box" approach can no longer meet the needs of manufacturers who need to iteratively upgrade their products quickly.
The intelligent selection of black sesame is open, not a "black box". Its software and hardware scheme can be decoupled, and the underlying software is completely provided to customers in the form of open source, which is customized by customers to realize the overall software integration.
Yang Yuxin thinks that black sesame can develop so fast as a local company, and A 1000Pro can greatly improve the computing performance, which is inseparable from their adherence to the principle of self-developed core IP (abbreviation of intellectual property).
There are two core IPS of black sesame. One is the image processor of automobile instrument liquid level.
NeuralIQ
ISP, simply speaking, is to switch between two scenes with too large light and dark gap, which may cause temporary blindness, and the human eye's perception performance is not very good in scenes such as entering and leaving the tunnel and the high beam in the opposite lane. The same is true for cameras, and the whole system can be "seen more clearly" after being processed by ISP module.
The other is the neural network module NPU, which is mainly used to identify the data sensed by sensors and analyze traffic participants such as vehicles and lane lines.
"Most companies with the most advanced autonomous driving technology will develop their own IP, such as NVIDIA, Tesla, Qualcomm, Mobileye and Huawei, all of which are independently developed IP." Yang Yuxin said.
"The window period is only 3 to 5 years, and it is impossible to catch up."
"We judge that in about three to five years, that is, in 2025, this window will slowly close. Because the automobile industry itself has relatively conservative characteristics, after the supplier system of new technologies is established, car companies will not give new manufacturers so many opportunities. "
Yang Yuxin described the urgency of chip companies' occupation.
At present, Black Sesame is establishing cooperative relations with major car companies, including China FAW, Bosch, Weilai, SAIC, BYD, Dongfeng, Zhongke Chuangda, Asia Pacific, Baolong, Sotorion, Newark Technology, Lianyou, etc. They have carried out a series of commercial cooperation on L2/3 ADAS and automatic driving perception system solutions.
Development history of black sesame intelligence
In mass production, Huashan II A 1000 chip will be used as an intelligent driving brain in Hongqi flagship SUV.
As for A 1000 Pro, which will be mass-produced by the end of 2022, Yang Yuxin did not disclose it.
Yang Yuxin said: "As long as you complete the mass production process with a head car company, it will greatly shorten the time to gain the trust of other car companies. So, our goal is to get the first few heads in our hands first. It may be 3~ 100 copies after completion, which is faster than the previous progress. "
At present, there are not many high-computing chips on the market, but they have reached a climax in models such as Weilai ET7, Ideal X0 1 (code name), SAIC Smart L7, and Extreme Fox Alpha S HI. These chips are from NVIDIA Orin and Huawei MDC8 10.
In addition to the above two products, Qualcomm Xiaolong cycling platform is planned to be mass-produced in 2022, and it is expected to cooperate with the models of Great Wall.
Everyone is vying for the opportunity to mass-produce L3 and L4 autopilot schemes. Therefore, for black sesame intelligence, mass production in 2022 will be a key sign of whether it can successfully occupy a place.
"The cost of self-research by car companies is high and it is unnecessary."
In the era of smart cars, the word "full-stack self-research" from the Internet industry has become a slogan that car companies are shouting more and more. Some people worry that assuming that the self-developed chip has also become a part of the full-stack self-developed chip of car companies, this will pose a threat to chip manufacturers such as Black Sesame Intelligence.
However, in Yang Yuxin's view, there is no need for car companies to learn from Tesla, and full-stack self-research is actually something that happened in a specific historical stage.
He took Apple's mobile phone as an example. "When Apple entered the mobile phone industry, it was an external circuit breaker. It is completely different from the traditional industrial chain, without the support of the industrial chain. If Apple wants to succeed, it must not only have good ideas, but also have strong technical strength, so Apple chose full-stack self-research. "
The same is true of Tesla's choice of full-stack self-research.
"The smart new energy vehicle defined by Tesla can be said to have become the direction pursued by all car companies. At this time, the whole industry moved. If car companies want to do similar things again, they don't need to do it themselves, because someone has already done it, and car companies can' get on the horse' at a faster speed and lower cost. "
Of course, this does not mean that car companies do not need to master the new capabilities of the supply chain.
On the contrary, Deng Kun believes that it is absolutely impossible for the main engine factory to participate in the formulation of the technical route, or even take the route into its own hands, simply provided by a first-class supplier.
Deng Kun, Vice President of Intelligent Application Engineering of Black Sesame
"In the future, maybe many automakers will become 0.5." Deng Wei said. In fact, at present, some car companies "pay tuition fees" through investment, and some car companies have done a lot of PoC (proof of concept) and rehearsal in the early stage.
In short, car companies are no longer the inherent form of traditional manufacturers.
Correspondingly, the form of a secondary supplier like Black Sesame Intelligence cannot remain unchanged. Deng Kun said that their customers may be traditional first-class suppliers or car companies, so the identity of black sesame will be divided between 0.5 and 1.5.
"Foreign giants in Nokia are optimistic about the development of domestic brands."
Nowadays, Contemporary Ampere Technology Co., Ltd., a well-known lithium battery supplier in China, was chosen by BMW to establish early technical cooperation in its fledgling and little-known early growth period, and then gradually embarked on the road of rapid development.
For chip companies, will conquering foreign brands directly make them expand faster in Zhang Zhilu?
Deng Jian didn't agree.
He revealed, "We have also had in-depth communication and exchanges with foreign brands, including BMW and Daimler Benz. Their development progress is not very fast, more like Nokia's position in the smart phone industry. "
We can't judge whether this assertion is true or not, but Black Sesame Intelligence is obviously more inclined to make good use of the domestic market policy orientation, cooperate with independent brands, quickly land products and increase market share.
Yang Yuxin further predicted that independent brands will surpass joint venture brands in a few years.
"The competitiveness of independent brands comes from the radical degree of the definition of selling points. The most typical example is that only China car factory dares to hand over the products with chip computing power 1000TOPS. Although they cannot be used in cars, they are a very attractive selling point for local users in China. " Yang Yuxin said.
The chip computing power of mass production cars in the domestic market has been improved year by year.
Scene clarity is another competitiveness of Yang Yuxin's own brand.
He analyzed and said, "10 years ago, China mobile phone brands let most foreign mobile phone brands go on in one fell swoop, relying on the sea tactics. Others/kloc-0 produce 5 models per year, and domestic brands may produce 50 models. Now the strategy of China automobile brand is very similar to that at that time, and it is being defined to more subdivided people, such as Great Wall Motor. "
Perhaps, as Yang Yuxin said, in a few years, independent brands will successfully counterattack.
However, in this process, the driving force and demand of independent brand car companies for intelligent innovation will definitely be stronger, which requires a group of powerful chip partners to help them realize the rapid transformation to automobile intelligence.
In the battle of automotive chips, it is certain that Black Sesame has a leading intelligent performance and is duty-bound. As for whether it can become the main force of automotive chips, we can only wait for time to reveal the answer.