# People who tried gpu-unlock on 3090s and such - does this make any difference on mixed-precision quants where you get some parts of the model quantized to <=Q8 but keep some tensors at BF16? — retained record

> Retained reddit record from reddit:LocalLLaMA.

Canonical: https://highsignal.app/data/records/95f4292016a22fd2

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


source: [reddit](https://highsignal.app/data/reddit) (reddit:LocalLLaMA)  published 2026-10-26 15:48 ingested 2026-10-27 02:39

[original source ↗](https://www.reddit.com/r/LocalLLaMA/comments/1xg0kle/people_who_tried_gpuunlock_on_3090s_and_such_does/)

## retained text



```
Dunno, curious if there's any info out there.
```



document fetched 2026-10-27 00:43 · canonical https://reddit.com/r/LocalLLaMA/comments/1xg0kle/people_who_tried_gpuunlock_on_3090s_and_such_does
