NVIDIA Introduces Three Computers for Autonomous Driving

Autonomous vehicles are no longer just futuristic robotic projects; today, they operate and transport passengers in the streets of many cities and countries. Companies like Waymo, WeRide, Zoox, and Apollo are among the most well-known.

Nvidia AV Computers

At CES 2025, NVIDIA CEO Jensen Huang highlighted the company’s growing presence in the autonomous vehicle sector, emphasizing its potential to become the first multi-trillion-dollar robotics industry.

NVIDIA strives to equip car manufacturers with the powerful software and hardware processing solutions necessary for developing cutting-edge autonomous driving systems. To achieve this, NVIDIA leverages three core computing platforms:

1. Training Platform: Powered by the high-performance DGX systems, NVIDIA facilitates the training of autonomous driving models using vast datasets, encompassing both real-world data and synthetically generated scenarios.
2. Omniverse platform: serves as a powerful platform for simulating and generating diverse driving scenarios. Within this platform, autonomous driving systems can be rigorously tested and validated through a wide range of scenarios. It also leverages NVIDIA’s Cosmos world model to ensure the generated data remains physically accurate and logically consistent, enriching the training process.