Parking System Function Description - Tigerwong

The parking lot system helps to achieve breakthrough AI performance. The T4 multi precision function can support all kinds of AI workloads with four different precision levels. Fp32 precision can provide 8.1 tflopfp16 precision can provide 65 tflopint8 precision can provide 130 topint4 precision can provide 260 top equipped with 2 t4gpu single servers can replace 54 CPU servers at most. For AI training, for AI reasoning workload. Two T4 single servers can replace nine dual socket CPU servers.The first batch of enterprises in China to start using T4 to expand and improve workload horizontal expansion include Baidu, Tencent, JD and iFLYTEK. China's leading computer manufacturer will also launch a series of T4 based servers. Including Inspur, Lenovo, Huawei, dawn, Inspur commercial machinery and Xinhua. Huang finally came up with a speech on the autopilot system platform nvidiadriveagx, which was applied to the new force manufacturers and freight companies. At the same time, it realizes the algorithm of processing 30 trillion learning per second. NVIDIA drive agxxavier platform is open and scalable.

NVIDIA's Chinese partners are developing autonomous driving technology from the last mile to the groundbreaking electric vehicle. Huang Renxun revealed that these Chinese partners include FAW, Xiaopeng automobile, singularity automobile, Tucson future, etc. These companies choose to equip future vehicles with high-performance and energy-efficient NVIDIA driveagx, and have established a strong partnership with NVIDIA for the future. Automatic driving is expected to be a great blessing for the logistics industry. In order to help accelerate its development, the parking lot system has driven the global freight demand and exacerbated the shortage of truck drivers with e-commerce. Chinese freight companies are adopting NVIDIA drive AgX to provide the computing performance required by driverless vehicles.

It is cooperating with plusai, a self driving start-up company, and fulltruck alliance cfta, a logistics company, to develop a fleet of driverless commercial trucks, FAW Group (fawgroup is one of China's largest truck manufacturers. It plans to make large-scale arrangements in 2021. These companies said they would use NVIDIA driveagxpegasu, an AI supercomputer with a computing power of 320 trillion times per second, to support the operation of their upcoming self driving trucks this week, and accelerate the upgrading of China's long-distance transportation industry. Zhijia technology is based on in-depth learning The automatic driving system can make use of the strong computing performance of nvidiadriveagxpegasu. Liu Wanqian, chief executive of Chi Chi technology, said in a statement: "the next generation of L4 automatic driving trucks of FAW will meet the strong demand of FTA transportation and logistics platform." in August, the pilot of automatic driving in San Jose City, California, autox, was launched in autox. A self driving distribution start-up company.

At the same time, it said that driveagxpegasu was selected for its last mile delivery vehicles. The company is using camera vision and other sensor inputs, combined with the redundancy and diversity of NVIDIA driveagx platform, to achieve safe delivery of driverless products. A number of new generation electric vehicle start-ups have emerged in China. These start-ups are committed to new development In the past few years, driveagxxavier has been selected by the new power enterprises of powertrain technology and automatic driving. NVIDIA driveagxxavier is being used to achieve this goal. It is planned that from 2020, Xiaopeng automobile, a start-up company headquartered in Guangzhou, China, and its main primary supplier Desai Xiwei will use people who can handle 30 trillion operations per second and have a power of only 30 watts Industrial intelligence computing platform.

Xiaopeng's new models realize L3 level Chinese automatic driving function in stages, and gradually realize full-automatic driving under various road conditions in the countryside, automatic search for parking spaces and parking in the parking lot, morning and evening peak congestion cruise, automatic switching of the best Lane on the expressway and adhering to continuous automatic driving, as well as personalized route customization based on the best navigation route Automatic driving. Xiaopeng car plans to build China's L3 class automatic driving system for mass production by 2021. It still needs manual control under certain conditions. L3 automatic driving enables the car to turn and brake itself. Xiaopeng automobile is fortunate enough to be the first Chinese car manufacturer to sell Xavier. He Xiaopeng, chairman of Xiaopeng automobile, said: the chip is to integrate smart cars with radical cars. Key components to distinguish.

Nvidial3 and L4 are pioneers of highly automatic driving processors. According to the joint statement issued at the signing ceremony between Xiaopeng automobile and NVIDIA, in addition to cooperation in hardware, the two sides will cooperate in more fields, including data acquisition and calibration, software development, and cloud based training. Xiaopeng automobile and global chip giants NVIDIA and NVIDIA China Partner DESAY West Wei signed the three party strategic cooperation agreement on the first day of the GTC world wide Conference on the world. Today, it jointly develops L3 level automatic driving technology that adaptable to China's traffic environment and driving scenes. Singularity car announces the development of the next generation of automatic driving vehicles with nvidiadriveagxxavierai chip, and driveagxxavier will also provide technology for the upcoming vehicles from Beijing's singularity car. Technical support.

Singularity vehicles in the parking lot system will use the excellent computing power and safety provided by NVIDIA automatic driving platform to develop L3 and higher-level automatic driving technology, and will officially complete the L3 technology research and development in 2019 and realize the mass production of L3 automatic driving in 2020. Singularity vehicle automatic driving team will use NVIDIA driveagxxavier platform to realize L3 and higher-level automatic driving Once the sensor is integrated on the vehicle and driveagxxavier platform, artificial intelligence and autopilot driver Dr. Huang Huang, chief scientist, will be able to upgrade the autopilot system easily through OTA without changing the hardware of the vehicle. By collecting data and learning from time to time, singularities can upgrade vehicle information and functions from time to time, and continue to provide users with information. Better personalized intelligent travel service. Singularity automobile can display a more open platform, apply more flexible expansion methods, realize more powerful computing power and reasonable power consumption in the intelligent driving of mass-produced vehicles in the future. Huang Yu said: use NVIDIA driveagxxavier platform. Regard driving safety as the core.

Rishidhall, vice president of NVIDIA Automotive business development, said: scalable nvid The iadriveagx platform will provide essential powerful computing power and more complicated AI technology for singularity car's safe automatic driving system. Singularity automobile company is visionary and is committed to improving intelligent automatic driving capability from time to time. Sfmotor plans to launch its first electric frequency divider sF5 next year, and says it will use driveagxxavier to develop the next generation autonomous system. Computing platform. In addition, sfmotor - driveagxxavier Zhijia technology, FAW and FTA for the next generation of autonomous computing platform - strategic partnership around autonomous trucks. Autox - driveagxpegasu is selected for the last mile of autonomous distribution vehicles to provide technical support for its next generation of autonomous distribution robots.

E-commerce giants JD and meituan will also choose NVIDIA jetsona Gxxavier builds the next generation of distribution robots. At present, JD and meituan have selected NVIDIA jetsonagxxavier platform. China accounts for more than 40% of the world's e-commerce transactions - about US $100 billion a year. According to the data of McKinsey Global Research Institute, in the past few decades, this figure has been higher than the sum of France, Germany, Japan, Britain and the United States. China is a leading e-commerce Service companies are developing artificial intelligence distribution robots to support this huge growth.They can transport goods between warehouses, speed up the last mile of distribution, and deliver products to consumers and enterprises. The next generation of distribution robots need to achieve large-scale computing performance in small packages. They include multiple high-resolution cameras and various sensors such as lidar We need to perceive the world around us, said Huang Renxun. We need to plan and drive exotic routes in a complex and dynamic rural environment. We need to identify and respond to pedestrians, cars, traffic lights, signs and other objects in real time.