Shifting Gear in Automotive AI: From Large Language Models to Foundation Models
A paradigm shift from software-defined vehicles to AI-defined vehicles is underway. By using less code, larger models, more compute, and bigger data, the industry is transitioning to the AI-centric approach of AV 2.0. We'll review how the three computers of the end-to-end NVIDIA DRIVE platform—with co-developed hardware, software, and infrastructure—is enabling the next generation of AI-defined vehicles. We'll also provide an overview of the fundamental differences and benefits that AI-defined vehicles provide over the previous generation.
- AI
- Automotive Software Expo
Speaker
NVIDIA Japan
DRIVE Solutions
Manager
Mate Szarvas
Currently working at NVIDIA leading the Japan DRIVE Solutions software team with a primary focus on supporting the transition of the Japanese automotive industry to the age of AI-first development. Except for a brief stay at BlackBerry QNX, career experience is centered on AI, such as developing large vocabulary speech recognition systems in academia and pioneering the use of convolutional neural networks for ADAS applications at DENSO.