AMD Instinct MI250X meets the minimum VRAM requirement for Q4 inference of ibm-granite/granite-3.3-2b-instruct. Review the quantization breakdown below to see how higher precision settings impact VRAM and throughput.
AMD Instinct MI250X can run ibm-granite/granite-3.3-2b-instruct with Q4 quantization. At approximately 583 tokens/second, you can expect Excellent speed - conversational response times under 1 second.
You have 127GB headroom, which is sufficient for system overhead and smooth operation.
| Quantization | VRAM needed | VRAM available | Estimated speed | Verdict |
|---|---|---|---|---|
| Q4 | 1GB | 128GB | 582.94 tok/s | ✅ Fits comfortably |
| Q8 | 2GB | 128GB | 385.25 tok/s | ✅ Fits comfortably |
| FP16 | 4GB | 128GB | 202.63 tok/s | ✅ Fits comfortably |