Chip scaling becomes a geopolitical project
Intel is positioning Moore's Law not as an industry inevitability but as a policy choice. Modern fabs cost tens of billions, AI compute demand is skyrocketing, and only a handful of companies can afford the race. Gelsinger's stance: without government support, the U.S. risks losing control of its future AI hardware stack.Why subsidies now define the hardware roadmap
Behind the rhetoric is a shift toward industrial policy-driven innovation.- AI training cycles require massive, consistent supply of advanced nodes, something commercial demand alone can't stabilize.
- The U.S. sees domestic semiconductor capacity as a buffer against geopolitical shocks.
- Intel needs predictable funding to compete with TSMC-backed and state-supported ecosystems.
Packaging, yield, and the next decade of compute
Intel is betting on advanced packaging as the bridge between slowing transistor scaling and explosive AI workloads. The pitch to Washington is simple: help fund capacity now to avert structural AI compute shortages in the late 2020s.
