Debate: will Anthropic really need only 500MW of inference next year?
zephyr_z9 · x · 2026-11-18
At TomorrowX, investor Brad Gerstner claimed Anthropic might need only 500MW of inference capacity next year. zephyrz9 pushes back: that assumes publicly served models keep getting smaller, but the observed trend is the opposite—models are getting bigger, so inference demand may not fall as projected.
More from Infra
- Banks kill ByteDance's $480M chip loan for Malaysia data center, closing the loophole — SumitGup · 2026-11-18
- LanceDB raises $57.5M Series B to rebuild the data stack for AI — ycombinator · 2026-11-18
- SGLang Summit Week Closes With a Sake Meetup in SF — DavidBennett__ · 2026-11-18
- Smuggled Blackwell fingerprints found in Chinese open-weight models; Jensen Huang faces scrutiny — Difficult-Limit-7551 · 2026-11-18
- First SGLang Summit draws 2,000+: v1.0 ships, consortium forms with NVIDIA, Google, Meta — aliscodes · 2026-11-18
- New ApexW optimizer claims lower loss than AdamW with half the VRAM of bitsandbytes — omendUUde · 2026-11-18