Musk: Chinese AI labs lead on performance per chip, a skill that outlasts shortages
victor_explore · x · 2026-09-27
Musk's take
Elon Musk says Chinese AI labs get by far the best performance per unit of compute, and guesses China solves its chip limits via lithography and domestic chipmaking in about two to three years — faster than most expect.
The quoter's extension
Chip-starved labs got good at doing more with less, and that skill persists once chips arrive. If your agent's margin depends on today's token prices, plan for them to drop.
Related event: Musk: China to Break Through Compute and Chip Bottlenecks in 2-3 Years(3 posts)→
More from Infra
- Mobile AI Factory: agentic pipeline builds on-device SLMs that beat 5-shot frontier models — kalyan_kpl · 2026-11-11
- Vivo's Starlink direct-to-cell service hits 10M users in two weeks across Brazil — XFreeze · 2026-11-11
- Nan Ya Plastics to spend NT$11.53B on fifth glass fabric plant on AI server demand — zephyr_z9 · 2026-11-11
- Three years building CPUs and spiking nets: one engineer's path to AI accelerators — k7agar · 2026-11-11
- CoreWeave: 2027 GPU renewals could add ~$4B annualized revenue, analyst skeptical — zephyr_z9 · 2026-11-11
- Running DeepSeek 4.1 locally on an $800 rig with a used RTX 3090 at 2-3 tok/s — LankyGuitar6528 · 2026-11-11