Intel Arc B580
GPU77
Run records
14
Models covered
4
Frameworks covered
L0 Self-reported × 1 · L1 Reproduced × 1 · L2 Cross-framework verified × 75
Evidence mix
Typical measured performance
Published benchmark averages per model × framework × quantization on this hardware.
- llama.cpp · IQ3_XS Decode 70.76 · Prefill 300.15 ·
- Ollama · IQ3_S Decode 65.12 · Prefill 253.41 ·
- Ollama · Q2_K Decode 60.73 · Prefill 314.93 ·
- llama.cpp · IQ2_M Decode 164.15 · Prefill 679.38 ·
- Ollama · Q3_K_L Decode 127.75 · Prefill 669.2 ·
- llama.cpp · IQ2_XS Decode 84.02 · Prefill 284.07 ·
- llama.cpp · IQ4_XS Decode 64.83 · Prefill 327.91 ·
- Ollama · IQ2_M Decode 78.85 · Prefill 313.28 ·
- Ollama · Q8_0 Decode 37.78 · Prefill 141.17 ·
- llama.cpp · IQ3_S Decode 138.57 · Prefill 710.44 ·
- llama.cpp · IQ3_XS Decode 131.95 · Prefill 488.13 ·
- Ollama · Q4_K_S Decode 101.11 · Prefill 350.9 ·
- Ollama · Q8_0 Decode 66.71 · Prefill 216.71 ·
- llama.cpp · IQ2_M Decode 59.78 · Prefill 260.58 ·
- Ollama · IQ1_S Decode 60.08 · Prefill 208.14 ·
- llama.cpp · IQ3_S Decode 58.55 · Prefill 274.06 ·
- llama.cpp · Q3_K_S Decode 65.46 · Prefill 319.04 ·
- Ollama · IQ2_XS Decode 70.97 · Prefill 264.22 ·
- llama.cpp · Q3_K_M Decode 57.84 · Prefill 315.13 ·
- Ollama · Q2_K Decode 67.02 · Prefill 263.91 ·
- llama.cpp · Q2_K Decode 169.62 · Prefill 813.51 ·
- llama.cpp · Q5_K_S Decode 120.67 · Prefill 490.78 ·
- Ollama · FP16 Decode 62.55 · Prefill 296.07 ·
- llama.cpp · Q3_K_S Decode 74.37 · Prefill 373.85 ·
- Strata · Q6_K Decode 60.78 · Prefill 199.46 ·
- Ollama · Q3_K_M Decode 110.96 · Prefill 517.6 ·
- Strata · Q5_K_M Decode 136.37 · Prefill 553.15 ·
- vLLM · IQ1_S Decode 66.31 · Prefill 217.11 ·
- vLLM · Q3_K_M Decode 56.47 · Prefill 307.9 ·
- Strata · IQ2_M Decode 88.38 · Prefill 340.69 ·
- Strata · Q3_K_M Decode 71.01 · Prefill 283.66 ·
- Ollama · IQ2_M Decode 67.87 · Prefill 259.33 ·
- llama.cpp · IQ3_S Decode 60.72 · Prefill 260.85 ·
- vLLM · IQ1_S Decode 85.09 · Prefill 376.72 ·
- vLLM · Q2_K Decode 76.78 · Prefill 352.63 ·
- vLLM · IQ3_XS Decode 157.7 · Prefill 718.89 ·
- vLLM · Q4_K_M Decode 135.38 · Prefill 477.15 ·
- Strata · FP16 Decode 92 · Prefill 368.95 ·
- Strata · Q4_K_M Decode 163.66 · Prefill 710.01 ·
- llama.cpp · IQ3_S Decode 47.04 · Prefill 172.07 ·
- llama.cpp · Q3_K_L Decode 43.5 · Prefill 143.66 ·
- Strata · IQ2_XS Decode 78.95 · Prefill 342.16 ·
- llama.cpp · IQ1_S Decode 98.24 · Prefill 506.66 ·
- llama.cpp · Q3_K_L Decode 70.52 · Prefill 317.93 ·
- vLLM · Q2_K Decode 89.29 · Prefill 348.31 ·
- Strata · IQ3_S Decode 101.32 · Prefill 490.59 ·
- llama.cpp · Q4_K_M Decode 117.19 · Prefill 568.87 ·
- vLLM · Q4_K_S Decode 137.77 · Prefill 674.05 ·
- llama.cpp · IQ2_XS Decode 61.39 · Prefill 302.91 ·
- llama.cpp · IQ3_XS Decode 56.63 · Prefill 277.66 ·
Fit check
Physics-based estimate from VRAM and memory bandwidth: pick a model to see which quantizations fit and the expected speed
Formula: tok/s = (bandwidth ÷ model size) × 0.55, adapted from llmfit; see run records for measured numbers
Key specs
- Vendor
- Intel
- VRAM
- 12 GB
- Memory bandwidth
- 456 GB/s
Run records
50 configurations
| Model | Framework | Quantization | Decode | Gen time | Samples |
|---|---|---|---|---|---|
| gemma-4-E4B-it | llama.cpp | IQ3_XS | 70.8 tok/s | — | 1 |
| gemma-4-E4B-it | Ollama | IQ3_S | 65.1 tok/s | — | 1 |
| Ministral-3-8B-Instruct-2512 | Ollama | Q2_K | 60.7 tok/s | — | 1 |
| iFlytek Spark-X2.5-4B | llama.cpp | IQ2_M | 164.2 tok/s | — | 1 |
| iFlytek Spark-X2.5-4B | Ollama | Q3_K_L | 127.8 tok/s | — | 1 |
| Qwen3.5-4B | llama.cpp | IQ2_XS | 84 tok/s | — | 1 |
| Qwen3.5-4B | llama.cpp | IQ4_XS | 64.8 tok/s | — | 1 |
| Qwen3.5-4B | Ollama | IQ2_M | 78.8 tok/s | — | 1 |
| Qwen3.5-4B | Ollama | Q8_0 | 37.8 tok/s | — | 1 |
| Qwen3.5-0.8B | llama.cpp | IQ3_S | 138.6 tok/s | — | 1 |
| Qwen3.5-0.8B | llama.cpp | IQ3_XS | 131.9 tok/s | — | 1 |
| Qwen3.5-0.8B | Ollama | Q4_K_S | 101.1 tok/s | — | 1 |
| Qwen3.5-0.8B | Ollama | Q8_0 | 66.7 tok/s | — | 1 |
| Qwen3.8-9B | llama.cpp | IQ2_M | 59.8 tok/s | — | 2 |
| Qwen3.8-9B | Ollama | IQ1_S | 60.1 tok/s | — | 1 |
| Llama-3.1-8B-Instruct | llama.cpp | IQ3_S | 58.5 tok/s | — | 1 |
| Llama-3.1-8B-Instruct | llama.cpp | Q3_K_S | 65.5 tok/s | — | 1 |
| Llama-3.1-8B-Instruct | Ollama | IQ2_XS | 71 tok/s | — | 1 |
| Llama-3-8B-Instruct | llama.cpp | Q3_K_M | 57.8 tok/s | — | 1 |
| Llama-3-8B-Instruct | Ollama | Q2_K | 67 tok/s | — | 1 |
| Llama-3.2-3B-Instruct | llama.cpp | Q2_K | 169.6 tok/s | — | 1 |
| Llama-3.2-3B-Instruct | llama.cpp | Q5_K_S | 120.7 tok/s | — | 1 |
| Llama-3.2-3B-Instruct | Ollama | FP16 | 62.5 tok/s | — | 1 |
| Qwen3.5-4B | llama.cpp | Q3_K_S | 74.4 tok/s | — | 1 |
| Qwen3.5-4B | Strata | Q6_K | 60.8 tok/s | — | 1 |
| Qwen3.5-0.8B | Ollama | Q3_K_M | 111 tok/s | — | 1 |
| Qwen3.5-0.8B | Strata | Q5_K_M | 136.4 tok/s | — | 1 |
| Qwen3.8-9B | vLLM | IQ1_S | 66.3 tok/s | — | 1 |
| Qwen3-8B | vLLM | Q3_K_M | 56.5 tok/s | — | 1 |
| Qwen3-8B | Strata | IQ2_M | 88.4 tok/s | — | 1 |
| Qwen3-8B | Strata | Q3_K_M | 71 tok/s | — | 1 |
| Llama-3.1-8B-Instruct | Ollama | IQ2_M | 67.9 tok/s | — | 1 |
| Llama-3-8B-Instruct | llama.cpp | IQ3_S | 60.7 tok/s | — | 1 |
| Llama-3-8B-Instruct | vLLM | IQ1_S | 85.1 tok/s | — | 1 |
| Llama-3-8B-Instruct | vLLM | Q2_K | 76.8 tok/s | — | 1 |
| Llama-3.2-3B-Instruct | vLLM | IQ3_XS | 157.7 tok/s | — | 1 |
| Llama-3.2-3B-Instruct | vLLM | Q4_K_M | 135.4 tok/s | — | 1 |
| Llama-3.2-3B-Instruct | Strata | FP16 | 92 tok/s | — | 1 |
| Llama-3.2-3B-Instruct | Strata | Q4_K_M | 163.7 tok/s | — | 1 |
| Kimi-K3 | llama.cpp | IQ3_S | 47 tok/s | — | 1 |
| Kimi-K3 | llama.cpp | Q3_K_L | 43.5 tok/s | — | 1 |
| Kimi-K3 | Strata | IQ2_XS | 79 tok/s | — | 1 |
| Qwen3.5-4B | llama.cpp | IQ1_S | 98.2 tok/s | — | 1 |
| Qwen3.5-4B | llama.cpp | Q3_K_L | 70.5 tok/s | — | 1 |
| Qwen3.5-4B | vLLM | Q2_K | 89.3 tok/s | — | 1 |
| Qwen3.5-4B | Strata | IQ3_S | 101.3 tok/s | — | 1 |
| Qwen3.5-0.8B | llama.cpp | Q4_K_M | 117.2 tok/s | — | 1 |
| Qwen3.5-0.8B | vLLM | Q4_K_S | 137.8 tok/s | — | 1 |
| Qwen3.8-9B | llama.cpp | IQ2_XS | 61.4 tok/s | — | 1 |
| Qwen3.8-9B | llama.cpp | IQ3_XS | 56.6 tok/s | — | 1 |