A16 PCIe
NVIDIA
16GB VRAM · 200 GB/s bandwidth · 4.5 FP16 TFLOPS · 250W TDP
The A16 PCIe has 16GB of VRAM with 200 GB/s memory bandwidth and 4.5 TFLOPS FP16 compute. At Q4 quantization, it can comfortably run Gemma 3 4B (62 tok/s), Qwen 2.5 7B (32 tok/s), Llama 3.1 8B (30 tok/s). Models larger than ~27B parameters won't fit even at Q4. Electricity cost is approximately $27/month at 250W TDP.
What LLMs can you run?
| Model | Params | Q4 Weight | Fit | Decode |
|---|---|---|---|---|
| Gemma 3 4B | 4.0B | 2 GB | comfortable | 62 tok/s |
| Qwen 2.5 7B | 7.6B | 4 GB | comfortable | 32 tok/s |
| Llama 3.1 8B | 8.0B | 4 GB | comfortable | 30 tok/s |
| Mistral Small 24B | 24.0B | 12 GB | tight | 10 tok/s |
| Gemma 3 27B | 27.4B | 14 GB | won't fit | |
| Qwen 2.5 Coder 32B | 32.5B | 16 GB | won't fit | |
| Llama 3.3 70B | 70.6B | 35 GB | won't fit | |
| Qwen 2.5 72B | 72.7B | 36 GB | won't fit | |
| Llama 3.1 405B | 405B | 202 GB | won't fit | |
| DeepSeek R1 671B | 671B | 336 GB | won't fit |
Similar GPUs
| GPU | VRAM | BW | TFLOPS | TDP |
|---|---|---|---|---|
| A2 | 16GB | 200 GB/s | 4.5 | 60W |
| A2 PCIe | 16GB | 200 GB/s | 4.5 | 60W |
| M1 Pro 16GB | 16GB | 200 GB/s | 10.6 | 30W |
| M2 Pro 16GB | 16GB | 200 GB/s | 13.6 | 30W |
| Arc Pro B50 | 16GB | 224 GB/s | 21.3 | 70W |
Compare with another GPU
Select another GPU to compare specs and model performance side by side.