# If Data Center permitting Slows, Throw Out Old A100 and Install Helios for 10,000x Improvement — retained record

> Retained reddit record from reddit:AMD_Stock.

Canonical: https://highsignal.app/data/records/60fd010d58d8073e

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:AMD_Stock)  published 2026-11-07 16:29 ingested 2026-11-08 02:38

[original source ↗](https://www.reddit.com/r/AMD_Stock/comments/1xqagir/if_data_center_permitting_slows_throw_out_old/)

## retained text



```
Old data centers from Meta and other hyperscalers with A100 (or equivalent generation compute) are in floor spaces of around \~**2 million square feet.**

Assuming **Meta deems Muse and Agentic AI too important** (a certainty), they could either just agree to whatever Hinojosa/Texans want (very likely), OR

**Throw out all the old compute from Nvidia's A100 and install AMD Helios**. The performance uplift will be around 10,000x. The sites are already zoned and permitted, and contracting with say MidAmerican Energy for more power is feasible.

Crazy that nobody is even calculating this. The election result may **actually trigger a faster and shorter upgrade cycle**. AMD may see surprise demand uplift from upgrading old hardware from 2028/2029. (2027 revenue is largely contracted.)
```



document fetched 2026-11-08 00:29 · canonical https://reddit.com/r/AMD_Stock/comments/1xqagir/if_data_center_permitting_slows_throw_out_old
