GPUs & Macs › NVIDIA L40S-8Q
NVIDIA L40S-8Q: local LLM speed
2 measured results for the NVIDIA L40S-8Q.
Measured results
| Model | Generate tok/s | Prompt tok/s | Quant | Backend | Test | Source |
|---|---|---|---|---|---|---|
| Llama 3.2 1B Instruct | 62.4 | 12,563 | Q4_K_M | llamafile | LocalScore avg | LocalScore details |
| Qwen2.5 14B Instruct | 3.5 | 343 | Q4_K_M | llamafile | LocalScore avg · part CPU | LocalScore details |
What fits on the NVIDIA L40S-8Q
Fit assumes the whole model in GPU memory with ~8K context.
| Model | Best quant that fits | File | Expected tok/s |
|---|---|---|---|
| gemma-4-12B-it | IQ4_XS | 6.4 GB | – estimate |
| Qwen3.5-9B | Q5_K_S | 6.4 GB | – estimate |
| Qwen3.5-4B | Q8_0 | 4.5 GB | – estimate |
| MiMo-V2.6-Distill-Qwen-9B | Q4_K_M | 5.8 GB | – estimate |
| gemma-4-E4B-it | Q5_K_M | 5.5 GB | – estimate |
| Qwen3-4B | Q8_0 | 4.3 GB | – estimate |
| gemma-3-1b-it | Q8_0 | 1.1 GB | – estimate |
| gemma-4-E2B-it | Q8_0 | 5.0 GB | – estimate |
| DeepSeek-R1-0528-Qwen3-8B | Q5_K_M | 5.9 GB | – estimate |
| Llama-3.2-3B-Instruct | Q8_0 | 3.4 GB | – estimate |
| Phi-4-mini-instruct | Q8_0 | 4.1 GB | – estimate |
| gemma-3-4b-it | Q8_0 | 4.1 GB | – estimate |
| Qwen3-14B | UD-IQ3_XXS | 6.0 GB | – estimate |
| gemma-3-12b-it | Q3_K_M | 6.0 GB | – estimate |
Similar hardware
NVIDIA RTX 2070 SUPER · NVIDIA RTX 4060 Ti · NVIDIA RTX 2060 SUPER · NVIDIA GTX 1070 Ti · NVIDIA RTX 3060 Ti · NVIDIA RTX 3070 · NVIDIA GTX 1070 · NVIDIA RTX 3070 Ti