CPU controlling the overall intelligence will be another way out for driverless cars

The birth of CPU has strengthened the development of intelligent industries such as computers, PCs, and mobile phones. From the earliest 4004 microprocessor in 1971, which laid the foundation for the development of personal computers and the intelligent functions of the system, to the Intel processor after 1989, which opened up the peak of personal computers, the computer has transformed from a command line to a point-and-click graphical operating environment.

The mobile phone CPU is like a computer, controlling the central system of the entire mobile phone. It runs the software in the memory and calls the database in the memory, and then analyzes and achieves the control purpose. Qualcomm, a household name, dominates the smartphone market. The newly released Snapdragon 820 is the top of the mobile phone CPU list, and only Apple A9X and Samsung 8890 can compare with it. MTK, NVIDIA, Intel, Huawei and other processors follow closely behind, occupying the vast majority of the low-end and mid-range markets.

Digging deeper, the intelligent hardware driven by the mobile Internet affects not only the Internet of Vehicles and the Internet of Things, but also a number of chip manufacturers behind the scenes. Can you believe that a car that can learn to drive can be trained by using ANI on the CPU?

Engineers from Nvidia have designed a system that can learn how to drive by observing the driver's driving behavior. This is a GPU-based system, and the main principle is that cameras and computers complement each other. Three cameras and two DRIVEPX computers are used to observe the driver's actions for 72 hours. The camera will capture and record the surrounding conditions and a large amount of data in 3D, and the GPU will track and store this information. This data will then be analyzed and decomposed into learning steps by the machine learning system Torch7, and finally create an autonomous driving system that does not require driver intervention. This may be another way out for driverless cars in the future.

This "training" has been conducted in multiple states in the United States, under different weather conditions and times, and on a variety of road types and road conditions. Test results show that the system can achieve 98 to 100% autonomous driving capabilities. Although the car can drive automatically, when the camera detects a situation that has not been encountered before, the system will remind the driver to take over the steering wheel and enter the learning mode. The content learned is then transmitted to the cloud in real time and integrated into future software upgrades, similar to learning robots such as AlphaGo, which can spontaneously learn and record responses to the same type of situation.

Digging deeper, the smart hardware driven by the mobile Internet not only affects the Internet of Vehicles and the Internet of Things, but also the chip manufacturers behind the scenes. So far, smart hardware has filled the global industry, and chip solutions have solved most of the "soft" problems, which has led to a resurgence in the research and development of smart devices in various industries, and smart products that can be connected to the Internet are everywhere. The most commonly used smart devices are car navigation and other daily-used smart devices. Among the many navigation-based ones, the customized navigation system of Meixing focuses on the interaction between the car and the computer, and properly completes the real synchronization between the two.

CPU has solved most of the "soft" problems in smart hardware, which has led to a rapid development of smart devices in various industries. Smart products that can be connected to the Internet are everywhere. The most commonly used smart devices are car navigation and other daily-use smart devices. Among the many navigation-based smart devices, the customized navigation system of Meixing focuses on the interaction between the car and the computer, and properly completes the real synchronization between the two.

But often the situation is rampant, with more advantages than disadvantages. Today's smart device chip solutions are generally the same, making the industry threshold lower and lower, and the applicable standards are also lowering. Small manufacturers attempt to end the future of their products in a fast-moving consumer-like way, just like a virus airborne and spreading throughout the Internet era, which is something we don't want to see.

Therefore, the top priority is to break the "non-intelligent" situation of smart hardware, so that future smart hardware will no longer be just an interactive and interconnected goose, and more fields and functions will be strengthened to the level of weak artificial intelligence.

As a winner of Toutiao's Qingyun Plan and Baijiahao's Bai+ Plan, the 2019 Baidu Digital Author of the Year, the Baijiahao's Most Popular Author in the Technology Field, the 2019 Sogou Technology and Culture Author, and the 2021 Baijiahao Quarterly Influential Creator, he has won many awards, including the 2013 Sohu Best Industry Media Person, the 2015 China New Media Entrepreneurship Competition Beijing Third Place, the 2015 Guangmang Experience Award, the 2015 China New Media Entrepreneurship Competition Finals Third Place, and the 2018 Baidu Dynamic Annual Powerful Celebrity.