AMD Radeon RX 7900 GRE + NVIDIA GTX 1080Ti
Desktop2
Run records
1
Models covered
1
Frameworks covered
L0 Self-reported × 2
Evidence mix
Typical measured performance
Published benchmark averages per model × framework × quantization on this hardware.
- llama.cpp · Q6_K Decode 12.53 · Prefill 116.21 ·
- llama.cpp · Q4_K_M Decode 11.13 · Prefill 71.46 ·
Key specs
- VRAM
- 27 GB
Run records
2 configurations
| Model | Framework | Quantization | Decode | Samples |
|---|---|---|---|---|
| Qwen3.8-27B | llama.cpp | Q6_K | 12.5 tok/s | 1 |
| Qwen3.8-27B | llama.cpp | Q4_K_M | 11.1 tok/s | 1 |