NVIDIA L40S meets the minimum VRAM requirement for Q4 inference of Qwen/Qwen3-Next-80B-A3B-Thinking-FP8. Review the quantization breakdown below to see how higher precision settings impact VRAM and throughput.
NVIDIA L40S can run Qwen/Qwen3-Next-80B-A3B-Thinking-FP8 with Q4 quantization. At approximately 32 tokens/second, you can expect Moderate speed - useful for batch processing.
You have 9GB headroom, which is sufficient for system overhead and smooth operation.
| Quantization | VRAM needed | VRAM available | Estimated speed | Verdict |
|---|---|---|---|---|
| Q4 | 39GB | 48GB | 32.02 tok/s | ✅ Fits comfortably |
| Q8 | 78GB | 48GB | 24.80 tok/s | ❌ Not recommended |
| FP16 | 156GB | 48GB | 13.21 tok/s | ❌ Not recommended |