GeForce RTX 3070 Ti
NVIDIA
8GB VRAM · 608 GB/s bandwidth · 21.8 FP16 TFLOPS · 290W TDP
The GeForce RTX 3070 Ti has 8GB of VRAM with 608 GB/s memory bandwidth and 21.8 TFLOPS FP16 compute. At Q4 quantization, it can comfortably run Gemma 3 4B (188 tok/s), Qwen 2.5 7B (99 tok/s), Llama 3.1 8B (93 tok/s). Models larger than ~14B parameters won't fit even at Q4. Electricity cost is approximately $31/month at 290W TDP.
What LLMs can you run?
| Model | Params | Q4 Weight | Fit | Decode |
|---|---|---|---|---|
| Gemma 3 4B | 4.0B | 2 GB | comfortable | 188 tok/s |
| Qwen 2.5 7B | 7.6B | 4 GB | comfortable | 99 tok/s |
| Llama 3.1 8B | 8.0B | 4 GB | comfortable | 93 tok/s |
| Mistral Small 24B | 24.0B | 12 GB | won't fit | |
| 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 |
|---|---|---|---|---|
| GeForce RTX 3060 Ti GDDR6X | 8GB | 608 GB/s | 16.2 | 225W |
| GeForce RTX 3070 Ti 8 GB GA102 | 8GB | 608 GB/s | 21.8 | 290W |
| Arc A580 | 8GB | 512 GB/s | 24.6 | 175W |
| Arc A750 | 8GB | 512 GB/s | 34.4 | 225W |
| Radeon Pro V520 | 8GB | 512 GB/s | 14.8 | 225W |
Compare with another GPU
Select another GPU to compare specs and model performance side by side.