Head to head

RTX 5090 vs RTX 5070 Ti

The RTX 5090 holds more — 32 GB against 16 GB — which decides what you can load at all. The RTX 5090 has more bandwidth at 1792 GB/s, which decides how fast tokens come out once a model fits.

Model (Q4_K_M, 8K)RTX 5090tok/sRTX 5070 Titok/s
Llama 3.1 8B Runs comfortably 222 Runs comfortably 113
Llama 3.1 70B Won't fit 2.65 Won't fit 1.35
Llama 3.2 3B Runs comfortably 478 Runs comfortably 249
Llama 3.2 1B Runs comfortably 1235 Runs comfortably 657
Llama 4 Scout 109B-A17B Won't fit 6.75 Won't fit 4.90
Qwen3 8B Runs comfortably 214 Runs comfortably 109
Qwen3 14B Runs comfortably 126 Runs comfortably 64.2
Qwen3 32B Runs comfortably 59.0 Won't fit 4.95
Qwen3 4B Runs comfortably 380 Runs comfortably 198
Qwen3 30B-A3B Runs comfortably 159 Won't fit 45.7
Qwen3 235B-A22B Won't fit 3.83 Won't fit 3.36
Qwen2.5-Coder 32B Runs comfortably 59.0 Won't fit 4.95
Qwen2.5 7B Runs comfortably 247 Runs comfortably 126
Qwen2.5 72B Won't fit 2.50 Won't fit 1.31
Gemma 3 4B Runs comfortably 405 Runs comfortably 211
Gemma 3 12B Runs comfortably 151 Runs comfortably 76.9

Specifications

RTX 5090RTX 5070 Ti
Memory32 GB16 GB
Bandwidth1792 GB/s896 GB/s
FP16 compute209 TFLOPS88 TFLOPS
Launch price$1,999$749
ArchitectureBlackwellBlackwell