Jensen Huang – TPU competition, why we should sell chips to China, & Nvidia’s supply chain moat
Summary
Jensen Huang argues that NVIDIA’s core value lies in transforming electrons to tokens, a complex process resistant to commoditization despite software valuation crashes. He defends CUDA’s dominance through its vast ecosystem and superior total cost of ownership, dismissing TPU competitors as niche solutions. Huang addresses supply chain bottlenecks as temporary and explains his late investments in Anthropic and OpenAI as necessary corrections to earlier strategic oversights regarding foundation model funding.
Topics discussed
Software valuations and the token economy
NVIDIA's ecosystem and the rise of AI agents
Supply chain alignment and manufacturing scale
TSMC capacity, HBM, and silicon photonics
Energy constraints and re-industrialization
NVIDIA vs. TPUs and ASICs: Architecture advantages
CUDA ecosystem and software moat
Hyperscaler dependencies and cloud strategy
Competitive landscape and performance benchmarks
ASIC development and hyperscaler investments
Supporting Neo-clouds and avoiding winner-picking
Pricing strategy and supply chain reliability
Vera Rubin and future performance leaps
AI security and the China compute debate
Chinese chip capabilities and open source models
National security implications of chip exports
Global ecosystem dominance and market dynamics
Moore's Law, segmentation, and NVIDIA's mission
Listen ad-free on Castria