# Nvidia demand is already declining as open models run on non-Nvidia chips

- **Speaker:** Greg Osuri
- **Attribution:** attributed
- **Said:** 2024-09-27
- **Theme:** GPU Economics — https://www.gregosuri.com/predictions/gpu-economics/
- **Source:** IOSG OFR 13th Panel \| GPU Symphony: Decentralized Compute Power (IOSG Ventures)
- **Watch at:** 00:21:49 — https://www.youtube.com/watch?v=_UG21_OaZLo&t=1309s
- **Canonical:** https://www.gregosuri.com/predictions/gpu-economics/2024-09-27-nvidia-demand-is-already-declining-as-open-models-b3jd/
- **Cite as:** Greg Osuri, "Nvidia demand is already declining as open models run on non-Nvidia chips," IOSG OFR 13th Panel \| GPU Symphony: Decentralized Compute Power (IOSG Ventures), 27 Sept 2024, 00:21:49. https://www.youtube.com/watch?v=_UG21_OaZLo&t=1309s

## Quote — verbatim

> We are already seeing a decline in Nvidia's demand, especially as we start cracking being able to deploy or infer on non-Nvidia chips... With MI300s from AMD you're able to run Llama 3.1 405B pretty well... there's a company called Exo Labs which seems very promising — they do cluster inference on even Macs. So as the models are getting open-sourced and as we see a lot of optimizations happening for non-Nvidia chips, I think the dominance of Nvidia is going to gradually go down.

## Context — site annotation, not spoken

Answering whether new GPU generations make decentralized GPU markets a race to the bottom. Captions garble "MI300" as "Mi 350s". Note the Exo Labs mention — consumer Macs doing cluster inference supports his local-compute thesis.