AMD Instinct MI250X meets the minimum VRAM requirement for Q4 inference of Qwen/Qwen3-235B-A22B. Review the quantization breakdown below to see how higher precision settings impact VRAM and throughput.
AMD Instinct MI250X can run Qwen/Qwen3-235B-A22B with Q4 quantization. At approximately 60 tokens/second, you can expect Good speed - acceptable for interactive use.
You have 13GB headroom, which is sufficient for system overhead and smooth operation.
| Quantization | VRAM needed | VRAM available | Estimated speed | Verdict |
|---|---|---|---|---|
| Q4 | 115GB | 128GB | 60.26 tok/s | ✅ Fits comfortably |
| Q8 | 230GB | 128GB | 38.25 tok/s | ❌ Not recommended |
| FP16 | 460GB | 128GB | 22.59 tok/s | ❌ Not recommended |