NVIDIA H100 PCIe 80GB meets the minimum VRAM requirement for Q4 inference of google/gemma-2b. Review the quantization breakdown below to see how higher precision settings impact VRAM and throughput.
NVIDIA H100 PCIe 80GB can run google/gemma-2b with Q4 quantization. At approximately 399 tokens/second, you can expect Excellent speed - conversational response times under 1 second.
You have 79GB headroom, which is sufficient for system overhead and smooth operation.
| Quantization | VRAM needed | VRAM available | Estimated speed | Verdict |
|---|---|---|---|---|
| Q4 | 1GB | 80GB | 398.90 tok/s | ✅ Fits comfortably |
| Q8 | 2GB | 80GB | 287.09 tok/s | ✅ Fits comfortably |
| FP16 | 4GB | 80GB | 135.74 tok/s | ✅ Fits comfortably |