CES 2025 Keynote by Gary Shapiro and Jensen Wang

Introduction

• Gary Shapiro: CEO and Vice Chairman of Consumer Technology Association
• Highlights potential of technology at CES
• Emphasis on tech transforming challenges into opportunities
• Tech uniting humanity, moving towards an autonomous future
• Breakthroughs in sustainability and global solutions

Introduction to Jensen Wang and NVIDIA

• Jensen Wang: Founder and CEO of NVIDIA
• Keynote highlights NVIDIA’s role in AI and accelerated computing
• Historical context: From early challenges to global influence
• NVIDIA’s innovations: Chatbots, robots, virtual worlds, etc.

Jensen Wang's Keynote

NVIDIA's Journey and Innovations

• 1993: NV1 launched, introducing game consoles in PCs
• 1999: Invention of the programmable GPU
• 2006: Introduction of CUDA for programmability of GPUs
• 2012: Rise of AI with AlexNet
• 2018: Google's Transformer (BERT) revolutionizing AI

AI and Machine Learning

• Deep impact on computing and applications
• AI revolutionizing every layer of tech stack
• Importance of modality in AI applications

GeForce and AI

• GeForce as a foundational tool for AI democratization
• Introduction of RTX Blackwell family
• Significant advancements in GPU technology and AI integration

NVIDIA Blackwell Architecture

• New GeForce RTX 50 series
  • 92 billion transistors, 4,000 TOPS, 4 petaflops AI
  • Enhanced ray tracing and shader capabilities
• Innovation in power and efficiency
• Application in laptops and desktops

AI Scaling Laws

• Pre-training: Large-scale data and compute needs
• Post-training: Reinforcement learning and feedback
• Test-time scaling: Dynamic computation allocation for AI tasks
• Growing demand for NVIDIA computing solutions

Blackwell AI Factory System

• Detailed construction and scalability
• Performance improvements in AI computation

Agentic AI and Enterprise Solutions

• Introduction of NVIDIA NIMS and NEMO for AI development
• Integration of AI into enterprise environments
• Development of specialized AI agents for various industries

Physical AI

• Introduction of NVIDIA Cosmos
  • World Foundation Model for understanding the physical world
• Connection to Omniverse for grounded AI simulation
• Applications in robotics and industrial AI

Industrial and Automotive Applications

• Industrial Digitalization: Partnership with Kion and Accenture
• Autonomous Vehicles: Partnership with various automotive companies
  • Introduction of Thor, a new processor for autonomous vehicles
• Importance of digital twins in manufacturing and automotive

Robotics and AI Future

• Humanoid robots as the next frontier
• NVIDIA Isaac Groot for robot development
• Synthetic motion generation and training AI models

Project Digits

• Introduction of a new compact AI supercomputer
• Based on NVIDIA's latest chip technologies
• Potential applications in everyday computing

Conclusion

• Reflection on NVIDIA’s progress and future prospects
• Emphasis on the growing integration of AI in various industries
• Thank you and video presentation for CES 2025