Webinar
NVIDIA Open Models for Physical AI
Let’s take a deep dive into NVIDIA’s physical AI open models.
Thursday, November 5
2–3 p.m. SGT | 5–6 p.m. AEDT
1 hour
Join us for NVIDIA Open Models for Physical AI to explore how intelligent systems perceive, reason, and act in the real world. This session breaks down NVIDIA’s physical AI stack—from the three-computer platform (NVIDIA DGX™, OVX™, and AGX™) to cutting-edge open models like NVIDIA Cosmos™ 3 for world simulation, GR00T for humanoid robotics, and NVIDIA Nemotron™ for reasoning and planning—equipping developers and leaders with a clear blueprint to build scalable embodied AI.
You’ll learn about:
- Fundamental components of NVIDIA's robotics ecosystem
- The three-computer platform for physical AI
- Cosmos, GR00T, and Nemotron for physical AI
Agenda:
NVIDIA's comprehensive Physical AI stack
Explore how open models are accelerating innovation across robotics, autonomous systems, and beyond.
Deep dive into NVIDIA's open models for Physical AI
- Cosmos 3: A next-generation world foundation model designed to generate physically accurate, photorealistic synthetic data and world simulations, dramatically reducing the cost and time of training embodied agents.
- GR00T: NVIDIA's foundation model purpose-built for humanoid robots that enables generalized robotic learning and dexterous manipulation through multimodal understanding.
- Nemotron: NVIDIA's family of open large language models optimized for enterprise reasoning, planning, and agentic workflows that serve as the cognitive backbone for physical AI systems.
Get Started With NVIDIA's Open Models—Additional Learning Resources
Q & A
