AMD Instinct MI300X meets the minimum VRAM requirement for Q4 inference of nari-labs/Dia2-2B. Review the quantization breakdown below to see how higher precision settings impact VRAM and throughput.
AMD Instinct MI300X can run nari-labs/Dia2-2B with Q4 quantization. At approximately 919 tokens/second, you can expect Excellent speed - conversational response times under 1 second.
You have 190GB headroom, which is sufficient for system overhead and smooth operation.
| Quantization | VRAM needed | VRAM available | Estimated speed | Verdict |
|---|---|---|---|---|
| Q4 | 2GB | 192GB | 919.40 tok/s | ✅ Fits comfortably |
| Q8 | 3GB | 192GB | 648.67 tok/s | ✅ Fits comfortably |
| FP16 | 5GB | 192GB | 321.13 tok/s | ✅ Fits comfortably |