Back to search results
Qwen3-8B · Q3_K_L
Intel Arc A770 16GB16GBllama.cpp v0.43L2 Cross-framework verified
This page aggregates 1 real-world runs of Qwen3-8B (Q3_K_L) on Intel Arc A770 16GB with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.
56.14 tok/s
Decode
Decode speed
200.64 tok/s
Prefill
Prefill speed
1.5 s
TTFT
Time to first token
14.2 GB
VRAM
VRAM usage
L2 Cross-framework verified
1
Measured runs
1
Independent sources
Other
Source platforms
2 days ago
Last verified
Performance
- llama.cpp · Q3_K_L (current)Decode 56.14 · Prefill 200.64 ·
- Ollama · Q5_K_S Decode 43.29 · Prefill 237.98 ·
- Strata · Q3_K_L Decode 82.65 · Prefill 307.54 ·
- Strata · Q5_K_S Decode 62.82 · Prefill 302.98 ·
Core figures
- Decode (avg)
- 56.14 tok/s
- Prefill (avg)
- 200.64 tok/s
- TTFT (avg)
- 1.5 s
- VRAM (avg)
- 14.2 GB
- MTP acceptance rate
- —
- TTFB
- 0.45 GB
- Power draw
- 207.4 W
Configuration
Member-level fields are taken from the most recent run
- Model
- Qwen3-8B
- Quantization
- Q3_K_L
- Framework
- llama.cpp
- Version
- v0.43
- Context length
- 4096 tokens
- Batch size
- 512
- GPU layers
- 20
- Flash Attention
- On
Hardware
Nominal and measured figures are shown side by side; whether it runs is the reader's call
- GPU
- Intel Arc A770 16GB
- Nominal VRAM
- 16 GB
- Measured VRAM (avg)
- 14.2 GB
- OS
- Windows 11 23H2
- Driver
- Intel Arc 31.0.101.5186
- CUDA
- 12.6
- Power draw
- 207.4 W
Sources & evidence
1 measured records in total, each traceable to its original source
v0.43 · Windows 11 23H2 · CUDA 12.6 · 4096 ctx
56.14 tok/s
Decode
200.64 tok/s
Prefill
1.5 s
TTFT
14.2 GB
VRAM
—
MTP
207.4 W
Power draw
L2 batch seed 2026-10-16: llama.cpp on hw_id=97 running model_id=53 at Q3_K_L