Nvidia announces new superprocessor for autonomous cars
At Nvidia’s developer conference GTC, company boss Jensen Huang announced the new “AI Data centre on Wheels”, which is designed for vehicles from the production year 2025. The DRIVE Atlan thus expands Nvidia’s roadmap for vehicle central computers. In addition to high computing power, it is supposed to be characterised by networking and security features at data centre level and also handle the complex AI algorithms that are indispensable for autonomous driving with ease.
In addition to the next-generation GPU architectures, the SoC integrates hardware accelerators for deep learning and computer vision; an arm architecture is responsible for general computing tasks. With this performance giant, Nvidia is targeting the software-defined cars of the future, which not only have to process huge amounts of code, but also receive continuous updates via the wireless interface. “The transportation industry needs a computing platform that it can rely on for decades. The software investment is too immense to repeat for each car,” Huang said.
The Atlan will integrate an Nvidia’s BlueField data processing unit (DPU), which delivers a broad range of advanced networking, storage and security services to support complex compute and AI workloads found in autonomous vehicles. BlueField offers full data-center-infrastructure-on-a-chip programmability, armed with a safe security enclave to prevent data breaches and cyberattacks. DRIVE Atlan is designed from the ground up to handle the large number of AI applications that run simultaneously in autonomous machines.
Daimler, Nvidia co-develop software-defined vehicle architecture
Since the company’s DRIVE Atlan, Orin and Xavier platforms are all programmable through open CUDA and TensorRT APIs and libraries, developers can leverage their investments across multiple product generations as they establish their future AV production roadmaps.
At the opportunity of the developers conference, Nvidia and Volvo Cars also announced to deepen their collaboration: The next generation of Volvo’s self-driving vehicles will use Nvidia’s DRIVE Orin platform as central computer. This SoC is offering a performance of 254 TOPS. The processor will run software developed in-house and by Zenseact, Volvo Cars’ autonomous driving software development company, and control among others safety-relevant vehicle systems such as braking and steering. The computing power and graphics processing delivered by the Orin enable advanced sensor suites needed for autonomous driving, such as the state-of-the-art lidar technology developed by Luminar, another of Volvo’s technology partners.
Volvo Cars’ SPA2 architecture will be available as hardware-ready for autonomous drive from production start. Its unsupervised autonomous driving feature, called Highway Pilot, will be activated when it is verified to be safe for individual geographic locations and conditions.
Related articles:
ARM adds functional safety IP to CPU and GPU architectures
NIO launches commercial selfdriving car with 1000TOPS
Daimler flaunts extra-wide hyperscreen for luxury EV
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
