▸ DEVICE UNDER TEST
NVIDIA Quadro RTX 3000 Max-Q — 6 GB VRAM.
▸ QUADRO RTX 3000 MAX-Q SPEC
- BRAND
- NVIDIA
- VRAM
- 6 GB GDDR6
- BANDWIDTH
- 288 GB/s
- FP16 COMPUTE
- 9.3 TFLOPS
- FP32 COMPUTE
- 4.7 TFLOPS
- CUDA CORES
- 1,920
- TENSOR CORES
- 240
- TDP
- 60 W
- ARCHITECTURE
- Turing
▸ AI CAPABILITY
194/ 428 models @ Q4
With 6 GB VRAM and 288 GB/s bandwidth, this GPU handles models up to 8B parameters.
Speed ≈ bandwidth / model_size × efficiency. A 7B model at Q4 runs at ~33 tok/s.
§ 01TOP MODELS FOR QUADRO RTX 3000 MAX-Q
194 FIT · SHOWING 20| MODEL | SIZE | VRAM Q4 | TOK/S | AVG |
|---|---|---|---|---|
| Aya Expanse 8B | 8B | 5.4 GB | 29 | 27.8 |
| Cogito 8B | 8B | 5.4 GB | 29 | 17.2 |
| DeepSeek R1 Distill Llama 8B | 8B | 5.4 GB | 29 | 36.2 |
| Gemma 3n E4B | 8B | 5.4 GB | 29 | 28.8 |
| Granite 3.3 8B | 8B | 5.4 GB | 29 | 24.6 |
| Llama-3.1-8B | 8B | 5.4 GB | 29 | 23.5 |
| Dolphin Llama 3 8B | 8B | 5.4 GB | 29 | 23.8 |
| Llama 3 8B | 8B | 5.4 GB | 29 | 37.3 |
| Tulu 3 8B | 8B | 5.4 GB | 29 | 31.5 |
| Ministral-8B | 8B | 5.4 GB | 29 | 19.3 |
| Nemotron-H 8B | 8B | 5.4 GB | 29 | 78.4 |
| Granite 8B | 8B | 5.4 GB | 29 | 26.1 |
| InternVL2 8B | 8B | 5.4 GB | 29 | 44.6 |
| MiniCPM-V 2.6 8B | 8B | 5.4 GB | 29 | 40.8 |
| RNJ-1 8B | 8B | 5.4 GB | 29 | 53.5 |
| Gemma 4 E4B | 8B | 5.4 GB | 29 | 32.1 |
| InternLM3 8B Instruct | 8B | 5.4 GB | 29 | 38.7 |
| Qwen3-VL 8B Instruct | 8B | 5.4 GB | 29 | 26.4 |
| Qwen3-Embedding 8B | 8B | 5.4 GB | 29 | — |
| Granite 4.1 8B | 8B | 5.4 GB | 29 | 25.1 |