NVIDIA DRIVE Hyperion
NVIDIA DRIVE Hyperion (officially renamed to “NVIDIA Hyperion” in September 2026) is NVIDIA's production-ready reference sensor-and-compute platform for robotaxis and Level 4 (L4)-capable autonomous vehicles.
It combines a standardized, pre-validated sensor suite (cameras, radar, lidar, ultrasonics and in-cabin sensors) with dual NVIDIA DRIVE AGX Thor system-on-a-chip (Blackwell-architecture-class). Each DRIVE AGX Thor SoC delivers up to 1000 INT8 TOPS and 2000 FP4 FLOPs.
Hyperion is paired with the NVIDIA Alpamayo software stack and the NVIDIA Halos safety layer. Redundant compute, sensors and functional-safety mechanisms meet ISO 26262 ASIL-D requirements, and the architecture scales from L2++ to Level 4. Adopters include Mercedes-Benz, Jaguar Land Rover, Lucid, Volvo, BYD, Geely, Isuzu and Nissan.
Classification
ADAS driving computer · serves as: Perception, Sensor fusion, AI Inference, High-level compute.
Which group NVIDIA DRIVE Hyperion belongs to and how it is built
Compute Modules is a subcategory of hardware components that provide processing power for robotic systems. It encompasses onboard computers, single-board computers (SBCs), AI accelerators, embedded processors, GPU/NPU compute modules, and other units responsible for processing sensor data and executing control logic. These modules form the foundation of modern autonomous, humanoid, and perception-capable robots.
An integrated device combining a compute module (SoC), cameras and a display. It runs the software stack responsible for road perception and driver-assistance control (lane centering, adaptive cruise control), communicating with the vehicle over the CAN bus.
Technical specification
Dual, redundant configuration
NVIDIA Blackwell-class GPU + ARM Neoverse V3AE CPU
Per single DRIVE AGX Thor SoC
Per single DRIVE AGX Thor SoC