GTC 2026 • 100-Minute Workshop
Deploy and Optimize LLMs and VLMs
on NVIDIA Jetson Thor
The definitive platform for Physical AI — perception, reasoning, and action at the edge.
Powered by Jetson
The next generation of edge AI platforms for generative AI workloads
New
Jetson Thor
- 2560-core Blackwell GPU
- 128GB Unified Memory
- Up to 2,070 TOPS (FP4) AI Performance
Production Ready
Jetson AGX Orin
- 2048-core Ampere GPU
- 32GB / 64GB Unified Memory
- Up to 275 TOPS AI Performance
Workshop Chapters
Setup and three hands-on chapters to master generative AI on Jetson Thor
~5 min Required
Setup Jetson
Find your device IP address and SSH into your Jetson Thor. Quick 5-minute setup.
1
~20 min Hands-On
Understanding LLMs and AI Microservices
Learn how to run and integrate AI microservices on NVIDIA Jetson for edge deployment.
2
~20 min Hands-On
Understanding VLMs
Explore real-time vision AI with the Live VLM WebUI on Jetson.
3
~20 min Hands-On
Voice + Vision with Multimodal AI Studio
Build and explore speech AI in voice pipelines with Multi-modal AI Studio on Jetson.