NVIDIA Teams Up with Global Giants to Lead Level 4 Autonomous Ecosystem

By /

NVIDIA recently announced a landmark collaboration with industry giants Uber, Stellantis, and Foxconn to pioneer the world’s largest commercially viable Level 4 autonomous vehicle fleet. The ambitious venture aims to launch testing in 2027, with mass production expected by 2028.

NVIDIA’s Autonomous Alliance: Uber, Stellantis, and Foxconn at the Helm

In a headline-grabbing move at the GTC conference, NVIDIA’s CEO Jensen Huang introduced the next-generation DRIVE AGX Hyperion 10 platform as the backbone of a global Level 4 autonomous vehicle ecosystem. This platform is set to empower a raft of automakers and ride-hailing apps, including Lucid, Mercedes-Benz, Aurora, Volvo, and Pony.ai, all of whom are engineering solutions compatible with Hyperion’s architecture.

Uber’s Drive Towards a Self-Driving Revolution

The collaboration hails a new era for ride-sharing giant Uber as it integrates NVIDIA’s Hyperion 10 into its ambitious plans to establish the world’s largest autonomous ride-hailing network. Targeting 2027 for deployment, the initial rollout aims for 100,000 autonomous vehicles on the road, equipped and trained via NVIDIA Cosmos for seamless data handling and AI evolution.

Stellantis Joins Forces to Develop Autonomous Platforms

Stellantis has cemented its role in the project’s future by developing the first Level 4 robotaxi platform, initially focusing on commercial and small passenger vehicles. These vehicles will leverage NVIDIA’s DRIVE AGX Hyperion 10 framework, promising comprehensive support for autonomous parking and unmanned driving modes, expecting initial deployment in the U.S. by 2028.

Foxconn: The Hardware Architects

Foxconn steps into the alliance as a cornerstone of hardware integration, ensuring high-performance computing (HPC) modules and sensor systems are seamlessly embedded within the vehicle platforms. Their expertise guarantees consistency and adherence to safety standards necessary for global deployment.

The Heart of NVIDIA’s Innovation: Hyperion 10 and DRIVE Thor

The cutting-edge DRIVE AGX Hyperion 10 is an innovation powerhouse, driven by the Blackwell architecture chipset, DRIVE AGX Thor. With each vehicle boasting up to 2,000 FP4 teraflops of computational muscle, capability is secured through:

  • 14 high-resolution cameras

  • 9 mmWave radars

  • 1 optical LiDAR

  • 12 ultrasonic sensors

This modular design enables swift adoption by various automotive manufacturers while maintaining an always-up-to-date status via online updates.

Pioneering Safe Autonomous Standards: The NVIDIA Halos Initiative

NVIDIA launches the groundbreaking “NVIDIA Halos” AI safety certification program, establishing pioneering industry standards to meticulously evaluate autonomous and robotic safety. Collaborations with major players like AUMOVIO, Bosch, Nuro, and Wayve underscore its commitment to ensuring all products meet rigorous physical AI safety and cybersecurity benchmarks.

Opening Doors to AI Training: A Rich Dataset Unfurled

To expedite the industry’s AI-driven advancements, NVIDIA has released a colossal, multimodal autonomous dataset from 25 countries, encapsulating 1,700 hours of crucial imagery, radar, and LiDAR data. This invaluable resource is designed to accelerate autonomous system validation and development under strict safety frameworks, utilizing platforms like Foretellix for real-world and simulated testing.

Related Posts

Scroll to Top