# swapped 5600x to 5600g and got 4x slower inference on vulkan — retained record

> Retained reddit record from reddit:LocalLLaMA.

Canonical: https://highsignal.app/data/records/c6a0c7a63a21c7d8

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


source: [reddit](https://highsignal.app/data/reddit) (reddit:LocalLLaMA)  published 2026-10-29 22:24 ingested 2026-10-30 02:40

[original source ↗](https://www.reddit.com/r/LocalLLaMA/comments/1xiy7of/swapped_5600x_to_5600g_and_got_4x_slower/)

## retained text



```
\- qwen/qwen3.8-27b fully in main GPU VRAM   
\- rx 7900 xtx 24gb  
\- ddr4 32gb 3200  
\- installed 5600g instead of 5600x to free some VRAM on main gpu  
\- disabled iGPU in LM Studio

got 15-20t\\s instead of 65-75. iGPU show 0% load. Main GPU - 100% load.   
ROCm performance didn't changed.  
What is wrong with Vulkan?
```



document fetched 2026-10-30 00:44 · canonical https://reddit.com/r/LocalLLaMA/comments/1xiy7of/swapped_5600x_to_5600g_and_got_4x_slower
