AMD Radeon RX 9070 XT vs NVIDIA GeForce RTX 5070 Ti
For running models locally, priced in India. At 8k context the first runs 14 models fully in memory, the second 16.
Specifications
| Spec | AMD Radeon RX 9070 XT | NVIDIA GeForce RTX 5070 Ti |
|---|---|---|
| GPU | Radeon RX 9070 XT | GeForce RTX 5070 Ti |
| Memory | 16 GB GDDR6 | 16 GB GDDR7 |
| Memory bus | 256-bit | 256-bit |
| Memory bandwidth | 640 GB/s | 896 GB/s |
| FP16 compute | 195 TFLOPS | 88 TFLOPS |
| Software | ROCm | CUDA |
| Board power | 304 W | 300 W |
| PCIe generation | Gen 5 | Gen 5 |
| PCIe lanes | x16 | x16 |
What each runs
Estimated decode speed with the most precise quant that fits each one est.
| Model | AMD Radeon RX 9070 XT | NVIDIA GeForce RTX 5070 Ti |
|---|---|---|
| Devstral Small 2 24B | offload | 35tok/sQ4_K_M |
| gpt-oss 20B | 80tok/sMXFP4 | 127tok/sMXFP4 |
| Gemma 4 12B | 27tok/sQ8_0 | 42tok/sQ8_0 |
| Qwen 3.5 9B | 36tok/sQ8_0 | 56tok/sQ8_0 |
| Mistral Small 3.2 24B | offload | 35tok/sQ4_K_M |
| Ministral 3 14B | 28tok/sQ6_K | 35tok/sQ8_0 |
| Qwen3 14B | 26tok/sQ6_K | 41tok/sQ6_K |
| DeepSeek R1 0528 Qwen3 8B | 36tok/sQ8_0 | 56tok/sQ8_0 |
| Qwen3 8B | 36tok/sQ8_0 | 56tok/sQ8_0 |
| Phi-4 14B | 26tok/sQ6_K | 41tok/sQ6_K |
| Qwen 3.5 4B | 73tok/sQ8_0 | 116tok/sQ8_0 |
| Ministral 3 8B | 35tok/sQ8_0 | 55tok/sQ8_0 |
| Granite 4.2 8B | 33tok/sQ8_0 | 52tok/sQ8_0 |
| Gemma 4 E4B | 42tok/sQ8_0 | 66tok/sQ8_0 |
| LFM2.5 8B-A1B | 144tok/sQ8_0 | 227tok/sQ8_0 |
| Llama 3.1 8B | 46tok/sQ6_K | 72tok/sQ6_K |
| Qwen 3.8 27B | offload | offload |
| Qwen 3.5 122B-A10B | offload | offload |
| Qwen 3.6 27B | offload | offload |
| Gemma 4 31B | offload | offload |
| Mistral Small 4 119B | offload | offload |
| Qwen 3.6 35B-A3B | offload | offload |
| Qwen 3.5 27B | offload | offload |
| Gemma 4 26B-A4B | offload | offload |
| Qwen 3.5 35B-A3B | offload | offload |
| GLM 4.7 Flash | offload | offload |
| Nemotron 3.5 Lightning 30B-A3B | offload | offload |
| Qwen3 Coder 30B-A3B | offload | offload |
| Kimi Linear 48B-A3B | offload | offload |
| Qwen3 32B | offload | offload |
| Gemma 3 27B | offload | offload |
| DeepSeek R1 Distill Qwen 32B | offload | offload |
| DeepSeek V4 Flash | won't run | won't run |
| GLM 5.3 Flash | won't run | won't run |
| Qwen 3.8 Flash Next | won't run | won't run |
| MiniMax M2.7 | won't run | won't run |
| Nemotron 3 Super 120B-A12B | won't run | won't run |
| Qwen3 Coder Next | won't run | won't run |
| gpt-oss 120B | won't run | won't run |
| GLM 4.5 Air | won't run | won't run |
| Qwen3 Next 80B-A3B | won't run | won't run |
| Llama 3.3 70B | won't run | won't run |