NVIDIA H200 SXM 141GB meets the minimum VRAM requirement for Q4 inference of vikhyatk/moondream2. Review the quantization breakdown below to see how higher precision settings impact VRAM and throughput.
NVIDIA H200 SXM 141GB can run vikhyatk/moondream2 with Q4 quantization. At approximately 727 tokens/second, you can expect Excellent speed - conversational response times under 1 second.
You have 137GB headroom, which is sufficient for system overhead and smooth operation.
| Quantization | VRAM needed | VRAM available | Estimated speed | Verdict |
|---|---|---|---|---|
| Q4 | 4GB | 141GB | 726.66 tok/s | ✅ Fits comfortably |
| Q8 | 7GB | 141GB | 508.71 tok/s | ✅ Fits comfortably |
| FP16 | 15GB | 141GB | 285.82 tok/s | ✅ Fits comfortably |