AMD Strix Halo (Radeon 8060S)
其他47
运行记录
15
覆盖模型
4
覆盖框架
L2 跨框架验证 × 47
证据分布
典型性能表现
同硬件上各模型 × 框架 × 量化的已上架实测均值。
- vLLM · Q8_0 Decode 72.19 · Prefill 307.07 ·
- Ollama · IQ3_XS Decode 123.41 · Prefill 403.1 ·
- llama.cpp · IQ1_S Decode 116.24 · Prefill 433.01 ·
- llama.cpp · Q5_K_S Decode 60.47 · Prefill 213.4 ·
- Strata · Q4_K_S Decode 105.67 · Prefill 508.86 ·
- Strata · Q6_K Decode 73.08 · Prefill 234.71 ·
- Ollama · IQ3_S Decode 74.72 · Prefill 343.92 ·
- Strata · IQ3_XS Decode 105.76 · Prefill 504.03 ·
- Strata · Q5_K_M Decode 70.97 · Prefill 388.51 ·
- llama.cpp · Q3_K_L Decode 205.24 · Prefill 943.68 ·
- llama.cpp · Q3_K_M Decode 229.54 · Prefill 1241.69 ·
- vLLM · IQ3_S Decode 249.57 · Prefill 1015.71 ·
- vLLM · Q3_K_L Decode 238.07 · Prefill 894.22 ·
- vLLM · Q3_K_L Decode 88.34 · Prefill 283.42 ·
- vLLM · Q8_0 Decode 54.8 · Prefill 273.05 ·
- Strata · IQ3_XS Decode 138.39 · Prefill 445.74 ·
- Strata · Q5_K_S Decode 95.58 · Prefill 525.15 ·
- vLLM · IQ1_S Decode 127.84 · Prefill 540.46 ·
- vLLM · Q4_K_S Decode 94.74 · Prefill 425.68 ·
- Ollama · Q5_K_M Decode 64.59 · Prefill 334.15 ·
- Ollama · Q8_0 Decode 48.28 · Prefill 205.87 ·
- llama.cpp · Q4_K_S Decode 106.73 · Prefill 450.35 ·
- Strata · Q4_K_S Decode 150.62 · Prefill 773.06 ·
- Strata · Q6_K Decode 108.81 · Prefill 551.07 ·
- llama.cpp · Q2_K Decode 188.51 · Prefill 863.98 ·
- llama.cpp · Q3_K_L Decode 168.83 · Prefill 829.48 ·
- vLLM · Q3_K_L Decode 174.98 · Prefill 636.17 ·
- llama.cpp · Q3_K_S Decode 117.04 · Prefill 404.22 ·
- llama.cpp · Q5_K_M Decode 77.99 · Prefill 423.17 ·
- Strata · IQ2_XS Decode 206.18 · Prefill 1005.37 ·
- Strata · IQ3_S Decode 166.26 · Prefill 669.87 ·
- llama.cpp · Q8_0 Decode 46.29 · Prefill 157.5 ·
- Ollama · Q2_K Decode 85.26 · Prefill 425.23 ·
- llama.cpp · FP16 Decode 124.48 · Prefill 656.36 ·
- llama.cpp · Q6_K Decode 171.39 · Prefill 894.49 ·
- vLLM · Q5_K_S Decode 228.54 · Prefill 950.08 ·
- llama.cpp · IQ4_XS Decode 127.7 · Prefill 503.71 ·
- Ollama · Q4_K_S Decode 115.67 · Prefill 506.03 ·
- Ollama · Q5_K_M Decode 87.89 · Prefill 401.06 ·
- vLLM · IQ2_XS Decode 355.91 · Prefill 1759.4 ·
- vLLM · IQ3_S Decode 267.01 · Prefill 1239.81 ·
- Strata · IQ3_S Decode 374.22 · Prefill 1937.23 ·
- Ollama · FP16 Decode 70.57 · Prefill 330.84 ·
- Strata · Q3_K_S Decode 228.54 · Prefill 966.84 ·
- Strata · Q6_K Decode 157.75 · Prefill 603.58 ·
- llama.cpp · IQ2_M Decode 142.24 · Prefill 506.53 ·
- Strata · Q8_0 Decode 92.35 · Prefill 312.7 ·
Fit 判定
按显存与内存带宽做物理估算:选择模型,查看各量化能否装下及预计速度
该硬件未收录显存数据,无法判定 Fit
估算公式:tok/s =(带宽 ÷ 模型占用)× 0.55,借鉴 llmfit;实测以运行记录为准
核心规格
- 厂商
- AMD
- 显存带宽
- 256 GB/s
运行记录
共 47 组配置
| 模型 | 框架 | 量化 | Decode | 生成耗时 | 样本数 |
|---|---|---|---|---|---|
| Qwen3.5-4B | vLLM | Q8_0 | 72.2 tok/s | — | 1 |
| Qwen3.5-4B | Ollama | IQ3_XS | 123.4 tok/s | — | 1 |
| Qwen3.5-9B | llama.cpp | IQ1_S | 116.2 tok/s | — | 1 |
| Qwen3.5-9B | llama.cpp | Q5_K_S | 60.5 tok/s | — | 1 |
| Qwen3.5-9B | Strata | Q4_K_S | 105.7 tok/s | — | 1 |
| Qwen3.5-9B | Strata | Q6_K | 73.1 tok/s | — | 1 |
| Qwen3.5-27B | Ollama | IQ3_S | 74.7 tok/s | — | 1 |
| Qwen3.5-27B | Strata | IQ3_XS | 105.8 tok/s | — | 1 |
| Qwen3.5-27B | Strata | Q5_K_M | 71 tok/s | — | 1 |
| Qwen3.5-0.8B | llama.cpp | Q3_K_L | 205.2 tok/s | — | 1 |
| Qwen3.5-0.8B | llama.cpp | Q3_K_M | 229.5 tok/s | — | 1 |
| Qwen3.5-0.8B | vLLM | IQ3_S | 249.6 tok/s | — | 1 |
| Qwen3.5-0.8B | vLLM | Q3_K_L | 238.1 tok/s | — | 1 |
| Qwen3.8-9B | vLLM | Q3_K_L | 88.3 tok/s | — | 1 |
| Qwen3.8-9B | vLLM | Q8_0 | 54.8 tok/s | — | 1 |
| Qwen3.8-9B | Strata | IQ3_XS | 138.4 tok/s | — | 1 |
| Qwen3.8-9B | Strata | Q5_K_S | 95.6 tok/s | — | 1 |
| Qwen3-8B | vLLM | IQ1_S | 127.8 tok/s | — | 1 |
| Qwen3-8B | vLLM | Q4_K_S | 94.7 tok/s | — | 1 |
| Qwen3-8B | Ollama | Q5_K_M | 64.6 tok/s | — | 1 |
| Qwen3-8B | Ollama | Q8_0 | 48.3 tok/s | — | 1 |
| Qwen3-4B | llama.cpp | Q4_K_S | 106.7 tok/s | — | 1 |
| Qwen3-4B | Strata | Q4_K_S | 150.6 tok/s | — | 1 |
| Qwen3-4B | Strata | Q6_K | 108.8 tok/s | — | 1 |
| Qwen3-1.7B | llama.cpp | Q2_K | 188.5 tok/s | — | 1 |
| Qwen3-1.7B | llama.cpp | Q3_K_L | 168.8 tok/s | — | 1 |
| Qwen3-1.7B | vLLM | Q3_K_L | 175 tok/s | — | 1 |
| Qwen3-4B-Thinking-2507 | llama.cpp | Q3_K_S | 117 tok/s | — | 1 |
| Qwen3-4B-Thinking-2507 | llama.cpp | Q5_K_M | 78 tok/s | — | 1 |
| Qwen3-4B-Thinking-2507 | Strata | IQ2_XS | 206.2 tok/s | — | 1 |
| Qwen3-4B-Thinking-2507 | Strata | IQ3_S | 166.3 tok/s | — | 1 |
| Qwen3-Coder-Next | llama.cpp | Q8_0 | 46.3 tok/s | — | 1 |
| Qwen3-Coder-Next | Ollama | Q2_K | 85.3 tok/s | — | 1 |
| Qwen2.5-0.5B-Instruct | llama.cpp | FP16 | 124.5 tok/s | — | 1 |
| Qwen2.5-0.5B-Instruct | llama.cpp | Q6_K | 171.4 tok/s | — | 1 |
| Qwen2.5-0.5B-Instruct | vLLM | Q5_K_S | 228.5 tok/s | — | 1 |
| Qwen2.5-Omni-3B | llama.cpp | IQ4_XS | 127.7 tok/s | — | 1 |
| Qwen2.5-Omni-3B | Ollama | Q4_K_S | 115.7 tok/s | — | 1 |
| Qwen2.5-Omni-3B | Ollama | Q5_K_M | 87.9 tok/s | — | 1 |
| Qwen3-Embedding-0.6B | vLLM | IQ2_XS | 355.9 tok/s | — | 1 |
| Qwen3-Embedding-0.6B | vLLM | IQ3_S | 267 tok/s | — | 1 |
| Qwen3-Embedding-0.6B | Strata | IQ3_S | 374.2 tok/s | — | 1 |
| gemma-4-E2B-it | Ollama | FP16 | 70.6 tok/s | — | 1 |
| gemma-4-E2B-it | Strata | Q3_K_S | 228.5 tok/s | — | 1 |
| gemma-4-E2B-it | Strata | Q6_K | 157.8 tok/s | — | 1 |
| gemma-4-E4B-it | llama.cpp | IQ2_M | 142.2 tok/s | — | 1 |
| gemma-4-E4B-it | Strata | Q8_0 | 92.3 tok/s | — | 1 |