llamaperf

Qwen3.8 27B

on RX 580 8GB

Tone: mixed
Sep 7, 2026
Throughput
7.4 t/s gen · 14.0 tokens/s prompt processing (prefill)Prompt processing measures input prefill speed; it does not tell you the time to first token.
Quant
Q3_K_M (GGUF)
System RAM
32 GB
VRAM reported
16 GB

Summary

User reports 7.39 t/s generation and 14 t/s prompt processing on 2x RX 580 8GB, 16 GB VRAM total, on an old DDR3 motherboard with 32 GB RAM. User calls the setup cheap but low speed.