RTX 5090 meets the minimum VRAM requirement for Q4 inference of ibm-granite/granite-3.3-8b-instruct. Review the quantization breakdown below to see how higher precision settings impact VRAM and throughput.
RTX 5090 can run ibm-granite/granite-3.3-8b-instruct with Q4 quantization. At approximately 295 tokens/second, you can expect Excellent speed - conversational response times under 1 second.
You have 28GB headroom, which is sufficient for system overhead and smooth operation.
| Quantization | VRAM needed | VRAM available | Estimated speed | Verdict |
|---|---|---|---|---|
| Q4 | 4GB | 32GB | 294.87 tok/s | ✅ Fits comfortably |
| Q8 | 9GB | 32GB | 228.70 tok/s | ✅ Fits comfortably |
| FP16 | 17GB | 32GB | 115.31 tok/s | ✅ Fits comfortably |