AMD Instinct MI210 meets the minimum VRAM requirement for Q4 inference of Qwen/Qwen2.5-72B-Instruct. Review the quantization breakdown below to see how higher precision settings impact VRAM and throughput.
AMD Instinct MI210 can run Qwen/Qwen2.5-72B-Instruct with Q4 quantization. At approximately 50 tokens/second, you can expect Good speed - acceptable for interactive use.
You have 28GB headroom, which is sufficient for system overhead and smooth operation.
| Quantization | VRAM needed | VRAM available | Estimated speed | Verdict |
|---|---|---|---|---|
| Q4 | 36GB | 64GB | 50.06 tok/s | ✅ Fits comfortably |
| Q8 | 71GB | 64GB | 35.70 tok/s | ❌ Not recommended |
| FP16 | 142GB | 64GB | 19.21 tok/s | ❌ Not recommended |