# Training Stable Diffusion: \~$600k, cheaper on Akash

- **Speaker:** Greg Osuri
- **Attribution:** attributed
- **Said:** 2022-11-03
- **Theme:** GPU Economics — https://www.gregosuri.com/predictions/gpu-economics/
- **Source:** Akash Weekly - November 2nd 2022 (Akash Network)
- **Watch at:** 00:34:57 — https://www.youtube.com/watch?v=XQVGt-fdKPY&t=2097s
- **Canonical:** https://www.gregosuri.com/predictions/gpu-economics/2022-11-03-training-stable-diffusion-600k-cheaper-on-akash-m6bh/
- **Cite as:** Greg Osuri, "Training Stable Diffusion: \~$600k, cheaper on Akash," Akash Weekly - November 2nd 2022 (Akash Network), 3 Nov 2022, 00:34:57. https://www.youtube.com/watch?v=XQVGt-fdKPY&t=2097s

## Quote — verbatim

> A protocol like Stable Diffusion, or a model — it costs about six hundred thousand dollars to get a basic training... with Akash we expect that to be, I don't know, four hundred thousand dollars, 400 to 500,000 — still a lot of money saving. So HPC makes a lot of sense for us... there's just enormous amounts of GPUs just sitting there idle.

## Context — site annotation, not spoken

AMA on target customers; adds that GPU users are new cloud users who care about performance and price, not traditional cloud features — a better fit for Akash.