NVIDIA builds redundancy and diversity into the AV platform to optimize occupant, pedestrian, and vehicle safety.
Safety is the most important aspect of a self-driving vehicle. By developing the end-to-end NVIDIA DRIVE™ platform that spans from the cloud to the car, we’re leading the way towards creating safer, more scalable, and more efficient transportation.
NVIDIA DRIVE is the world’s first scalable AI platform that spans the entire range of autonomous driving. The platform consists of hardware, software, and infrastructure that work together to enable the production of automated and self-driving vehicles. With high-performance computing, the vehicle can understand in real time what’s happening around it, precisely localize itself on a high-definition map, and plan a safe path forward.
NVIDIA DRIVE Infrastructure consists of the hardware, software, and simulation environments required for the massive data collection, deep learning development, and testing to support and train autonomous vehicles.
This AI infrastructure helps developers create and quickly train DNN models to enable highly accurate perception systems for autonomous vehicles. These include dozens of neural networks that separately cover perception of lanes and road boundaries, road markings, signs, vehicles, wait conditions, free space, vulnerable road users, and more. This makes DRIVE Infrastructure the ideal solution for the entire transportation industry.
Before autonomous vehicles can safely navigate the road, engineers must first test and validate the exaflops of data that enable the vehicle to drive itself, enabling a truly redundant testing methodology. Powered by the NVIDIA DGX SuperPOD™ system with NVIDIA DGX A100 or DGX H100, our data center solutions enable the testing of millions of permutations to increase model accuracy for high levels of safety.
But there’s no feasible way to physically road test all these various weather, lighting, or road conditions. The ability to test in a realistic simulation environment is essential to providing safe self-driving vehicles. NVIDIA DRIVE Sim™, built on NVIDIA Omniverse™, solves for this by providing a physically accurate platform designed for safe, fast, and efficient AV testing and validation at scale.
Safety is our highest priority at every step of the research, development, and deployment process. It begins with a methodology that emphasizes diversity and redundancy in the design, validation, verification, and lifetime support of the autonomous system.
Designed to support L2-L5 of the SAEJ3016 specification, the DRIVE AGX™ architecture enables vehicle manufacturers to build and deploy self-driving vehicles that are functionally safe and are compliant to international safety standards. These include ISO 26262 and ISO/DIS 21448, NHTSA recommendations, and global NCAP requirements.
NVIDIA Privacy Policy
NVIDIA DRIVE is an open AV development platform for software developers worldwide. It includes DRIVE Software plug-ins that allow partners to integrate their technology with the NVIDIA DRIVE AGX system.