NVIDIA DRIVE - Safety

The next generation of transportation is autonomous. NVIDIA is leading the way, driven by a mission to develop self-driving technology that enables safer, less congested roads and mobility for all. Find out more about our vision, innovations and processes with our NHTSA-based NVIDIA Self-Driving Car Safety Report submitted to the National Highway Traffic Safety Adminstration (NHTSA).

The Four Pillars of Safe Autonomous Driving

Safe autonomous driving is built on four fundamental pillars. With high-performance compute at their core, these tenets illustrate NVIDIA’s dedication to safety and ensure a robust self-driving technology development cycle.


Pillar 1: An Artificial Intelligence Design and Implementation Platform

A safe AI driver needs a compute platform that spans the entire range of autonomous driving—from assisted highway driving to robotaxis. This platform needs to combine deep learning, sensor fusion and surround vision to enable the car to make split-second decisions based on massive amounts of data.

To safely operate, self-driving vehicles require supercomputers powerful enough to process all the sensor data in real time. Our underlying hardware solutions include:

For self-driving cars, processing performance translates to safety. The more compute, the more sophisticated the algorithm, the more layers in a deep neural network (DNN), and the greater number of simultaneous DNNs that can be run simultaneously. You can find an overview of our autonomous driving safety software policy here:

  • Planning a Safer Path: Mathematically Proven and Validated in Simulation, NVIDIA Safety Force Field Protects Against Real-World Traffic

Pillar 2: Development Infrastructure That Supports Deep Learning

In addition to in-vehicle supercomputing hardware, NVIDIA solutions power the data centers used to solve critical challenges faced in the development of safe AVs. A single test vehicle can generate petabytes of data each year. Capturing, managing and processing this massive amount of data for not just one car, but a fleet requires an entirely new computing architecture and infrastructure.


Pillar 3: Data Center Solution for Robust Simulation and Testing

The ability to test in a realistic simulation environment is essential to providing safe self-driving vehicles. By coupling actual road miles with simulated miles in a high-performance data center solution, manufacturers can comprehensively test and validate their technology.


Pillar 4: Best-in-Class, Pervasive Safety Program

Self-driving technology development must follow a pervasive safety methodology that emphasizes diversity and redundancy in the design, validation, verification, and lifetime support of the entire autonomous system. These programs should follow recommendations from federal and international agencies such as the National Highway Traffic Safety Administration, International Organization for Standardization, and the global New Car Assessment Program.


NVIDIA DRIVE™ OS, is the first functional safety (FuSa) operating system designed specifically for accelerated computing and artificial intelligence. This foundational software stack for autonomous vehicles consists of an embedded real-time operating system (RTOS), NVIDIA hypervisor, NVIDIA® CUDA® libraries, NVIDIA TensorRT™ and other modules that provide access to the hardware engines.

DRIVE OS offers a safe and secure execution environment with services such as secure boot, security, firewall and over-the-air (OTA) updates. Quality of service is achieved with the embedded RTOS and NVIDIA hypervisor.

DRIVE OS meets the requirements of a robust operating system for autonomous vehicle software, including:

  • Safety: Automotive industry safety standards, such as ASPICE and ISO 26262
  • Security: Comprehensive security model
  • Performance: High-performance compute acceleration engines (CUDA, TensorRT, NvMedia, NvStreams)
  • Scalability: Scaling from advanced driver assistance to full self driving
  • Production: A broad ecosystem of industry partners and experience

Safety-certified Compute APIs, including CUDA, TensorRT and NvMedia with high-speed transport (NvStreams) enable deep neural networks (DNNs) running on the vehicle to process sensor inputs at high speeds. By running these DNNs simultaneously, vehicles can process data from a variety of sensors in real-time for Level 4 and Level 5 autonomous driving capabilities.

Three-layer safety supervision (3LSS) is a safety framework that monitors overall system integrity from the Tegra SoC to the safety MCU, providing a scalable, safety-assured channel for handling error conditions and making system-level decisions.

QNX is safety-certified and proven in the automotive industry. It’s an ISO 26262 ASIL D-certified RTOS that shares a common heritage with Linux, providing an easy development path for most companies.

NVIDIA’s virtualization technology is the automotive industry’s most advanced, enabling strict resource isolation, protecting hardware and memory, as well as quality of service, ensuring the integrity of data as it’s processed through the system, all with near-native performance and very low overhead.