In-Vehicle Computing for AI-Defined Cars

Performance for Next-Generation Vehicles

The NVIDIA DRIVE AGX™ platform includes all the hardware and software necessary to develop automated driving functions and immersive in-cabin experiences. It’s an open and modular platform running the NVIDIA DriveOS™ SDK, and when coupled with supported sensors and accessories, enables manufacturers to build autonomous driving functions and in-vehicle AI applications. These can then be further enhanced over the air (OTA).

Hardware

DRIVE Orin

The NVIDIA DRIVE Orin™ SoC (system-on-a-chip) delivers 254 TOPS (trillion operations per second) and is the central computer for intelligent vehicles. It’s the ideal solution for powering autonomous driving capabilities, confidence views, digital clusters, and AI cockpits. The DRIVE Orin product family lets developers build, scale, and leverage one development investment across an entire fleet, from Level 2+ systems all the way to Level 5 fully autonomous vehicles.

DRIVE Thor

The DRIVE Thor SoC is our next-generation centralized car computer, combining advanced driver assistance and an AI cockpit on a single safe and secure system. This AV processor uses our latest CPU and GPU advances—including the NVIDIA Blackwell GPU architecture for transformer and generative AI capabilities. DRIVE Thor features 8-bit floating point support (FP8)—to deliver an unprecedented 1,000 INT8 TOPS/1,000 FP8 TFLOPS/500 FP16 TFLOPS of performance while reducing overall system cost.

DRIVE Hyperion

The NVIDIA DRIVE Hyperion™ reference architecture accelerates development, testing, and validation by integrating DRIVE Orin-based AI compute with a complete sensor suite. It includes 12 exterior cameras, three interior cameras, nine radars, one interior radar, 12 ultrasonics, one front-facing lidar, and one lidar for ground-truth data collection.

Software

DriveOS

The foundation of the DRIVE software stack, DriveOS is the first safe operating system for in-vehicle accelerated computing. It includes NVIDIA® CUDA® libraries for efficient parallel computing implementations, NVIDIA TensorRT™ for real-time AI inference, and NvMedia for sensor input processing.

DriveWorks

NVIDIA DriveWorks provides middleware functions on top of DriveOS that are fundamental to autonomous vehicle development. These consist of the sensor abstraction layer (SAL) and sensor plug-ins, data recorder, vehicle I/O support, and a deep neural network (DNN) framework. It’s modular, open, and designed to be compliant with automotive industry software standards.

DRIVE Chauffeur

NVIDIA DRIVE Chauffeur is an AI-assisted driving platform that can handle both highway and urban traffic with the utmost safety. It can use the high-performance compute and sensor set of NVIDIA DRIVE Hyperion to drive from address to address. For those who want to drive, the system also provides active safety features and the ability to intervene in dangerous scenarios.

DRIVE Concierge

NVIDIA DRIVE Concierge delivers new, always-on intelligent services for the driver and passengers. It uses NVIDIA Avatar Cloud Engine (ACE) to serve as everyone’s digital assistant, making recommendations, helping book reservations, making phone calls, accessing vehicle controls, and providing alerts using natural language. DRIVE Concierge also features a confidence view into what DRIVE Chauffeur sees around the car and what it’s planning next. It even serves as a valet, automatically parking and summoning the car.

Find out how you can start developing AI-powered autonomous vehicles.

  翻译: