Smart cars are growing so fast

 CNMOULDING began to produce automotive injection molded parts for TESLA matching

Today I will share some of my views on Smart Cars

In the smart car just started today, the more you start, the more you buy new and not old.

There is two logic behind this view:

The first point is that in the process has not reached the bottleneck today, chip performance is growing exponentially.

Nvidia's latest generation of intelligent chip Orin adopts 7 nm production technology, which can realize the operation performance of 200 TOPS per second, which is 7 times higher than that of the previous generation Xavier system-level chip.

Then NIVIDA Guanxuan ALTAN automatic chip, a single SOC of 1000 is more than four times the ORIN, dozens of times the Xavier.

Before and after the production time is only a few years, but the performance difference is several times, in the future more and more complex, the more intelligent cars will rely on the on-chip performance, once the bottleneck of computing power, the more advanced intelligent services are basically declared out of the way.

The second point is that smart cars are iterative quickly, from traditional cars to digital cockpit only in the stage of primary evolution, and even now smart cars are far from smart, just adding more screens and providing some local content. This is just the normal operation of a smartphone.

The imagination space of intelligent cars in the future is very large, especially the human demand has not been fully developed, the competition based on software and architecture will become increasingly hot, and the evolution of intelligent cars in the future will be beyond everyone's imagination.

Tesla is one of the pioneers of smart cars, changing the way the auto industry plays. The operating logic of selling software based on hardware proves that software will play a greater role in the future than hardware.

Tesla focused on autopilot and digital cockpit at the beginning of its establishment.

Tesla autopilot development quite twists and turns, first cooperate with the Moblieye, to the choice NIVIDA, finally, choose the whole stack self-research, the development belongs to their own autopilot. Tesla may have difficulties, but independent R & D is, after all, the safest choice.

Another surprise is that the interior was completely reformed when the Model S was launched, almost abandoning all keys, retaining only a large screen and opening the concept of a digital cockpit.

Later, whether Wei Lai, Xiaopeng, or Xiaomi are basically along these two roads.

But I have always believed in the phrase "there are no identical people in the world ". Everyone's different experiences, education, emotions, personality make everyone unique, and the result is that everyone decorates different houses, reads different information, has different mobile phone interfaces, and should have different cars.

Once the individual is transplanted into the car, it involves not only sitting posture, air conditioning temperature, but also driving habits-fierce or gentle, music preferences-soothing or excited, and even as a second space for movement, the car has different people's different preferences, can be satisfied and so on.

Before Cadillac introduced the new pure electric vehicle LYRIQ, there was a vision that the skylight glass was divided into four areas, and the corresponding four passengers could adjust the light transmission separately.

Earlier Bose an attempt was made to install sound on different car seats, and the music played could only be heard by specific people. Along this line of thinking, can different people also watch different videos in the same car?

Therefore, smart cars in the future have a lot of imagination, is only just beginning.

Although many cars are developing autopilot and digital cockpit, the two are relatively independent and unrelated, such as Wei Lai Automobile NOP autopilot auxiliary system and digital cockpit in data communication. The screen is only used as a self-driving display device, each updated independently.

In the second half of the car industry, more and more big Internet factories, such as Baidu, such as Ali, such as Horizon, have been trying to get through the two parts, although they divide autonomous driving and intelligent cockpit into two parts: outside and inside, and finally form "man-machine co-driving ".

A debate remains over whether smart driving and smart cockpit should merge as if horizon had organized a debate on both sides before the shanghai auto show.

If we look to the future, integration is the trend of the times.

We believe that when the technical barriers of the traditional automobile —— the engine, gearbox, chassis —— disappear, the power source of the pure electric vehicle is replaced by the motor and the battery, the logic of driving the automobile forward is changed from hardware to software, and all hardware is defined by software, which can truly realize car ".

Unmanned driving driven by software, digital cockpit content ecology supported by software and service system supported by cloud service. As a result, the operating system, computing platform, AI algorithm and sensing equipment of the technology enterprise completely push back the old three parts, reshape the automobile ecology, and make the intelligent operating system become the second important engine of the automobile.

Once all hardware is connected to the cloud through software, it is natural to achieve personal customization, which must be based on the integration of intelligent driving and intelligent cockpit.

Self-learning and evolutionary ability will become the core difference between the new generation of the intelligent cockpit and the traditional generation of the intelligent cockpit.

The traditional generation of intelligent cockpit focuses on big data collection, not on individuals. Wei Lai Automobile NOMI provides intelligent voice service. After collecting all the data, analyzing and processing in the cloud and enriching the database, the instructions are more and more accurate, but they can not do thousands of people.

Zebra Zhihang put forward a new point, different people ask the same question, the answer should be different. For example, the owner is in a bad mood, sends out the navigation demand to the voice assistant, the voice assistant will take the initiative to make the judgment to provide a soothing route, for example, the owner is in a hurry. Voice assistants offer a seemingly blocked route that takes the least time after calculation.

This is essentially intelligent interaction.