Dwarkesh visits Intel: AI compute grows 3x/year but fab tools only allow ~2x production
Dwarkesh Podcast · rss · 2026-10-22
Dwarkesh Patel spent two days at Intel talking to everyone from the head of process development to the CTO about what chips will look like in 2045. The core tension: AI compute demand has grown 3x a year for a decade, but the world only has enough chipmaking tools to roughly double leading-edge production. With animated explainers, the episode covers gate-all-around transistors, backside power, CFET's decade-long path to shipping, hybrid bonding and chip-to-chip bottlenecks, why High-NA EUV pays off for Intel but not TSMC, replacing copper with ruthenium, Intel's 10nm lessons, whether Intel can catch TSMC, and the Terafab rethink. Flashcards and a full transcript included.
More from Infra
- Liquid AI Launches RIPL: Ultra-Low-Latency Inference Engine With Sub-50ms End-to-End Responses — JosephJacks_ · 2026-10-23
- OpenJet claims Qwen3.8-27B coding runs 1.9x faster than llama.cpp on one RTX 3090 — JeremyCMorgan · 2026-10-23
- Jensen Huang lays out NVIDIA's full Vera Rubin agentic system: NVL72, MGX, CMX and RTX — rohanpaul_ai · 2026-10-23
- ONNX to Transformers.js: running models directly in your browser — JosephJacks_ · 2026-10-23
- MiniMax M3.1 Flash free for 3 days on 0G, with 1M context and 512K output — MiniMax_AI · 2026-10-23
- "We bet on Cloudflare and it paid off": R2 and D1 at billion scale — dinasaur_404 · 2026-10-23