Ex-Meta AI exec: a 10T MoE likely burns ~1000 J/token, not the 5 J claimed in human-brain energy comparison
NaveenGRao · x · 2026-10-12
David Holz estimated a frontier 10T MoE on GB300 NVL72 uses 5 J/token (50 W at human thinking speed), close to the brain's 20 W. Naveen Rao pushes back: the real figure is likely closer to 1000 J/token. He adds that if motor control counts as a task, animals do real-time control on milliwatts, so tokens aren't comparable to human thought in quality anyway.
Related event: Midjourney Founder Estimates AI and Brain Thinking Energy on Par(3 posts)→
More from Infra
- Yandex reports data center damage causing service outages, 'Internyet' meme ensues — kingslyj · 2026-10-12
- 4070 Ti Super beats RTX 5070 2x on MiniMax H3 int8 in local inference benchmark — smellyelon · 2026-10-12
- Strata Runs a Big Open Model Locally on a Normal Gaming PC — alex_verem · 2026-10-12
- Why keep making LLMs bigger? A case for DeepSeek and Qwen's small-model extremism — gnukeith · 2026-10-12
- WSJ: Inside Silicon Valley's compute wars — Google paid $2.7B for Shazeer, then lost him to OpenAI — dotey · 2026-10-12
- Qualcomm AI uses entropy and CKA to pick layers, beating EAGLE3 defaults in speculative decoding — Qualcomm-AI-Research · 2026-10-12