# One GPU per prompt — 7 billion AI users implies staggering GPU demand

- **Speaker:** Greg Osuri
- **Attribution:** attributed
- **Said:** 2026-02-12
- **Theme:** GPU Economics — https://www.gregosuri.com/predictions/gpu-economics/
- **Source:** "DACM Insights: Decentralizing AI, The Akash Approach" (DACM Insights)
- **Watch at:** 00:13:55 — https://www.youtube.com/watch?v=BQONg2f0zQc&t=835s
- **Canonical:** https://www.gregosuri.com/predictions/gpu-economics/2026-02-12-one-gpu-per-prompt-7-billion-ai-users-implies-stag-clyb/
- **Cite as:** Greg Osuri, "One GPU per prompt — 7 billion AI users implies staggering GPU demand," "DACM Insights: Decentralizing AI, The Akash Approach" (DACM Insights), 12 Feb 2026, 00:13:55. https://www.youtube.com/watch?v=BQONg2f0zQc&t=835s

## Quote — verbatim

> Inference — in order to infer a model, you need one GPU per one prompt. Simply put, every prompt you send... takes 10 seconds to return the answer; that 10 seconds is being computed on one GPU... So if you have 7 billion people in the world, if everybody's on AI, you can imagine how many GPUs we need.

## Context — site annotation, not spoken

Answering why AI needs so much compute/energy, as setup for the decentralization argument.