Loading content...
Loading GPU benchmarks...
AMD Instinct MI300X meets the minimum VRAM requirement for Q4 inference of rednote-hilab/dots.ocr. Review the quantization breakdown below to see how higher precision settings impact VRAM and throughput.
| Quantization | VRAM needed | VRAM available | Estimated speed | Verdict |
|---|---|---|---|---|
| Q4 | 4GB | 192GB | 784.33 tok/s | ✅ Fits comfortably |
| Q8 | 7GB | 192GB | 512.91 tok/s | ✅ Fits comfortably |
| FP16 | 15GB | 192GB | 271.55 tok/s | ✅ Fits comfortably |