AMD Strix Halo 128GB vs Tesla P40 24GB
For running local LLMs · 40 reports across 8 models
Highest reported generation speed by model family
These maxima come from different community setups. Model size, quantization, GPU count, offloading and concurrency may differ. They do not establish which GPU is faster for the same workload. Check the comparison conditions →
| Model | AMD Strix Halo 128GB | Tesla P40 24GB |
|---|---|---|
| Qwen3.8up to 2400B | 153.3n=19 | 91.0n=3 |
| Qwen3.6up to 35B | 77.9n=9 | — |
| DeepSeek V4 Flashup to 285B | 35.0n=4 | — |
| GLM-4.7up to 30B | 40.0n=1 | — |
| LFM2.5up to 3B | 113.0n=1 | — |
| Nex-N2.5-mini | 76.9n=1 | — |
| Qwen3-Coder-Next | 36.8n=1 | — |
| Qwen3.5up to 122B | 49.0n=1 | — |