NVIDIA Physical AI workshop

Schedule (Day 1)

• Check-in (9:30-10am)

• Opening Introduction of the event

• Isaac Sim/Lab Crash Course

• Break-1 (11:00-11:20am)

• Turbo-Charging LLMs with TensorRT-LLM + RAG (Demo)

• Lab-1 (Genesis: A Generative and Universal Physics Engine for Robotics)

• Lunch (1:30-2:15pm) (Provided at the premises)

• Video Search Summarization

• Lab-2 (VLM Video Summarization)

• Connect with NVIDIA Experts

• Break-3 (4:30-5:00pm)

• Wrap

Focus / Objective (Day 1)

• Check-in: Device & repo sanity (Jetson/Laptop)

• Opening Introduction of the event: Physical AI mindset: Sim → Edge

• Isaac Sim/Lab Crash Course: Scene graph, sensors, logging

• Turbo-Charging LLMs with TensorRT-LLM + RAG (Demo): TRT-LLM optimization + RAG on Jetson

• Lab-1: Build a digital twin + sensor stack; log synthetic data

• Video Search Summarization: Schema, ingest, retrieval, summaries

• Lab-2: Search/summarize warehouse/traffic video

• Connect with NVIDIA Experts: Drop-in guidance (Sim/VSS/ROS)

• Wrap: Capture blockers; assign Day-2 goals

---

Schedule(Day 2)

• Check In (9:30-10am)

• Jetson Thor Overview

• Lab-3 (Edge Deploy)

• 1:1 Mentorship Blocks

• Lunch + Partner Corner

• Vision of NVIDIA’s Physical AI Community Interaction

• Demos

Focus / Objective (Day 2)

• Check In: Helpdesk

• Jetson Thor Overview: Setup, perf checklist, telemetry

• Lab-3: DeepStream/Isaac ROS container → Jetson

• 1:1 Mentorship Blocks: 30-min slots: throughput, memory, ops

• Lunch + Partner Corner: Tooling & ops show-and-tell

• Vision of NVIDIA’s Physical AI Community Interaction: POC sign-ups, office hours, community

• Demos: Futuristic solutions showcase