Quadro K1100M holds 48 of the models in our catalogue and is, in practice, a Q4_K_M card — the largest it takes is LFM3-VL 2B at Q4_K_M.
Neighbours in memory rather than in price: memory decides whether a card can do the job at all, so two cards of the same size at different prices is the comparison you are making. Ordered by memory, then bandwidth — not by the value column, which is worked out from bandwidth and price alone and therefore rewards a cheap card whatever its software stack does to that bandwidth in practice. Read it as one input, not as a ranking.
NVIDIA Quadro K1100M — 2 GB VRAM.
- Brand
- NVIDIA
- VRAM
- 2 GB GDDR5
- Bandwidth
- 45 GB/s
- FP16 Compute
- 0.5 TFLOPS
- FP32 Compute
- 0.5 TFLOPS
- CUDA Cores
- 384
- TDP
- 45 W
- Architecture
- Kepler
With 2 GB VRAM and 45 GB/s bandwidth, this GPU handles models up to 1.61B parameters.
Speed ≈ bandwidth / model_size × efficiency. A 7B model at Q4 runs at ~5 tok/s.
| Model | Size | VRAM Q4 | tok/s | Avg |
|---|---|---|---|---|
| GPT-2 XL 1.5B | 1.61B | 1.5 GB | 22 | 5.1 |
| stablelm-2-1_6b | 1.6B | 1.5 GB | 23 | 9.5 |
| Falcon-H1 1.5B | 1.55B | 1.4 GB | 23 | 43.8 |
| Qwen2.5-Coder-1.5B | 1.5B | 1.4 GB | 24 | 19.6 |
| Qwen2 Math 1.5B | 1.5B | 1.4 GB | 24 | 19.6 |
| Qwen 2.5 1.5B | 1.5B | 1.4 GB | 24 | 30.2 |
| Yi Coder 1.5B | 1.5B | 1.4 GB | 24 | 14.6 |
| Stella en 1.5B v5 | 1.5B | 1.4 GB | 24 | 52.4 |
| Phi-1 1.3B | 1.42B | 1.4 GB | 25 | 7.2 |
| Phi-1.5 1.3B | 1.42B | 1.4 GB | 25 | 7.2 |
| DeepSeek Coder 1.3B | 1.35B | 1.3 GB | 27 | 16.8 |
| EXAONE-4.0-1.2B | 1.3B | 1.3 GB | 28 | 22.6 |
| OPT 1.3B | 1.3B | 1.3 GB | 28 | 5.3 |
| MiniCPM-V 4.6 | 1.3B | 1.3 GB | 28 | 28.5 |
| LFM2.5-1.2B-Thinking | 1.2B | 1.2 GB | 30 | 22.3 |
| Llama-3.2-1B | 1.2B | 1.2 GB | 30 | 10.1 |
| LFM2 1.2B | 1.2B | 1.2 GB | 30 | 15.6 |
| Zamba2 1.2B | 1.2B | 1.2 GB | 30 | 41.5 |
| TinyLlama 1.1B | 1.1B | 1.2 GB | 33 | 13.6 |
| MiniCPM5 1B | 1.08B | 1.1 GB | 33 | 30.8 |