NVIDIA RTX 6000 Ada meets the minimum VRAM requirement for Q4 inference of ai-forever/ruGPT-3.5-13B. Review the quantization breakdown below to see how higher precision settings impact VRAM and throughput.
NVIDIA RTX 6000 Ada can run ai-forever/ruGPT-3.5-13B with Q4 quantization. At approximately 139 tokens/second, you can expect Excellent speed - conversational response times under 1 second.
You have 41GB headroom, which is sufficient for system overhead and smooth operation.
| Quantization | VRAM needed | VRAM available | Estimated speed | Verdict |
|---|---|---|---|---|
| Q4 | 7GB | 48GB | 138.58 tok/s | ✅ Fits comfortably |
| Q8 | 13GB | 48GB | 92.38 tok/s | ✅ Fits comfortably |
| FP16 | 27GB | 48GB | 48.20 tok/s | ✅ Fits comfortably |