AMD Instinct MI300X meets the minimum VRAM requirement for Q4 inference of meta-llama/Llama-Guard-3-1B. Review the quantization breakdown below to see how higher precision settings impact VRAM and throughput.
AMD Instinct MI300X can run meta-llama/Llama-Guard-3-1B with Q4 quantization. At approximately 858 tokens/second, you can expect Excellent speed - conversational response times under 1 second.
You have 191GB headroom, which is sufficient for system overhead and smooth operation.
| Quantization | VRAM needed | VRAM available | Estimated speed | Verdict |
|---|---|---|---|---|
| Q4 | 1GB | 192GB | 858.24 tok/s | ✅ Fits comfortably |
| Q8 | 1GB | 192GB | 665.07 tok/s | ✅ Fits comfortably |
| FP16 | 2GB | 192GB | 342.68 tok/s | ✅ Fits comfortably |