# q38fn on mixed gpus, which engine? — retained record

> Retained reddit record from reddit:LocalLLaMA.

Canonical: https://highsignal.app/data/records/f43d3eb5a646f7d4

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


source: [reddit](https://highsignal.app/data/reddit) (reddit:LocalLLaMA)  published 2026-10-30 12:16 ingested 2026-10-31 02:40

[original source ↗](https://www.reddit.com/r/LocalLLaMA/comments/1xjdrs7/q38fn_on_mixed_gpus_which_engine/)

## retained text



```
So i am running 2 radeon vii and a rtx 3080. Currently i am using strata which does not support mixed backends, so my 3080 is just expensive swap space.

Is there any inference engine that supports fast q38fn over mixed backends like rocm/cuda/vulkan but is also fast? Llama.cpp is way too slow sadly.
```



document fetched 2026-10-31 00:45 · canonical https://reddit.com/r/LocalLLaMA/comments/1xjdrs7/q38fn_on_mixed_gpus_which_engine
