Milk Road AI How Power and Edge AI Are Redefining the U.S.-China AI Competition w/ Patrick & Duncan
Dec 12, 2025
Patrick and Duncan dive into the crucial role of AI infrastructure over mere models in the U.S.-China AI race. They discuss China's innovative tactics to close its chip gap while emphasizing the importance of power and data center development for AI dominance. Edge computing gains traction as AI in devices like drones and autonomous vehicles increases. Investment strategies emerge around data centers and emerging edge chips, accentuating the wide-ranging implications of energy needs and technological advancements for future growth.
AI Snips
Chapters
Transcript
Episode notes
Chip Exports Are Geopolitical Strategy
- Selling older chips to China can be a strategic play to keep them a generation behind rather than pushing them to self-sufficiency.
- China responds by accelerating domestic chip development and restricting H200 imports unless justified, shrinking the gap.
System Scale Beats Single-Chip Efficiency
- China can match Western raw compute by clustering more, less-efficient chips into larger systems at much higher power cost.
- Their advantage is system-level scale and faster data-center builds, not single-chip efficiency.
Power Is The New Strategic Resource
- Power and data-center build speed are emerging as decisive factors in the U.S.–China AI race.
- China's rapid gigawatt additions and faster construction materially shorten the competitiveness gap.