Nvidia GPU At the Technology Conference (GTC) 2021, The Provided by the company A new system-chip for autonomous cars. Its name is Drive Atlon and it is a processor with next generation GPU configuration which will be implemented in 2025 models.
Similarly, Drive Atlon comes with new ARM CPU cores. As well as the Bluefield data processing unit, which helps complete the chip workload. In particular, processes related to artificial intelligence required to operate the automation systems of cars. In this regard, Nvidia CEO Zhen-Hoon Huang said:
“Our new Atlon is truly a technological marvel, combining Nvidia’s all the strengths of Bluefield’s artificial intelligence, automotive, robotics, security and secure data centers to provide a safe, self – driving fleet. The car is too big to do it again. ”
Drive Atlon can run more than 1,000 billion operations per second
Atlan’s technology will allow it to perform more than 1,000 trillion operations per second. About three times as much as Orin can do, the chip that comes with the company’s vehicles in 2022.
In fact, the 2025 models will come with more power and the possibility of upgrading through software updates via OTA (Over the Air, for its abbreviation in English). To do this, Adlon will use the “same APIs, CUDA and TensorRT libraries” used on existing drive chips.
Volvo will use homogeneity in its next vehicles
In other announcements, Nvidia said Drive Oren will be used in cars coming from the Swedish brand Volvo. In fact, it will be offered in next year’s XC90 model. Its function is to guarantee the data processing of the Lidar and Vision sensor.
That function completes the main functions of the car, namely power management and driver assistance. Together they guarantee a good performance of the vehicle.
Nvidia has introduced a new Hyperion platform
In addition to the Drive Atlas, the company introduced the Drive Hyperion 8. It is an update aimed at its “AI Supercomputer on Wheels” that drives the development of autonomous vehicles with Level 4 or higher automation.
It includes two SoC DRIVE Orins responsible for processing “data from 12 external cameras, three internal cameras, nine radars and a litar sensor”.
With these announcements, Nvidia clarifies its intention to improve the driving system of its vehicles, while at the same time paving the way for autonomy. At this point, they can only wait to see how much they will upgrade their cars.
The Great Wall plans to launch a hydrogen-powered vehicle