NVIDIA RTX PRO 4000 Blackwell vs AMD Radeon RX 7900 XTX
For running models locally, priced in the UK. At 64k context the first runs 48 models fully in memory, the second 47.
In short
Both have 24 GB of memory for models; the NVIDIA RTX PRO 4000 Blackwell has 672 GB/s of bandwidth and the AMD Radeon RX 7900 XTX 960 GB/s. At 64k context the NVIDIA RTX PRO 4000 Blackwell runs 48 of the models we track fully in memory and the AMD Radeon RX 7900 XTX 47; only the NVIDIA RTX PRO 4000 Blackwell holds Nemotron 3.5 Lightning 30B-A3B. On the models both run fully in memory, the AMD Radeon RX 7900 XTX is about 1.3× as fast as the NVIDIA RTX PRO 4000 Blackwell (estimated, median across models).
Updated . Speeds are estimates; how they are worked out.
Specifications
| Spec | NVIDIA RTX PRO 4000 Blackwell | AMD Radeon RX 7900 XTX |
|---|---|---|
| GPU | RTX PRO 4000 Blackwell | Radeon RX 7900 XTX |
| Memory | 24 GB GDDR7 | 24 GB GDDR6 |
| Memory bus | 192-bit | 384-bit |
| Memory bandwidth | 672 GB/s | 960 GB/s |
| FP16 compute | 46.9 TFLOPS | 123 TFLOPS |
| Software | CUDA | ROCm |
| Board power | 145 W | 355 W |
| PCIe generation | Gen 5 | Gen 4 |
| PCIe lanes | x16 | x16 |
What each runs
Estimated decode speed with the most precise quant that fits each one est.
| Model | NVIDIA RTX PRO 4000 Blackwell | AMD Radeon RX 7900 XTX |
|---|---|---|
| Qwen 3.8 27B | 19tok/sQ4_K_M | 24tok/sQ4_K_M |
| Qwen 3.6 27B | 17tok/sQ5_K_M | 25tok/sQ4_K_M |
| Qwen 3.6 35B-A3B | 77tok/sQ4_K_M | 97tok/sUD-Q4_K_M |
| Qwen 3.5 27B | 20tok/sQ4_K_M | 25tok/sQ4_K_M |
| Gemma 4 26B-A4B | 52tok/sUD-Q5_K_M | 66tok/sUD-Q5_K_M |
| Qwen 3.5 35B-A3B | 77tok/sQ4_K_M | 98tok/sQ4_K_M |
| GLM 4.7 Flash | 38tok/sQ4_K_M | 48tok/sQ4_K_M |
| Nemotron 3.5 Lightning 30B-A3B | 94tok/sMXFP4 | offload |
| gpt-oss 20B | 62tok/sMXFP4 | 78tok/sMXFP4 |
| Gemma 4 12B | 30tok/sQ8_0 | 38tok/sQ8_0 |
| Qwen 3.5 9B | 36tok/sQ8_0 | 45tok/sQ8_0 |
| Gemma 3 27B | 19tok/sQ4_K_M | 24tok/sQ4_K_M |
| Ministral 3 14B | 19tok/sQ6_K | 24tok/sQ6_K |
| Qwen3 14B | 18tok/sQ6_K | 23tok/sQ6_K |
| DeepSeek R1 0528 Qwen3 8B | 23tok/sQ8_0 | 29tok/sQ8_0 |
| Qwen3 8B | 23tok/sQ8_0 | 29tok/sQ8_0 |
| Phi-4 14B | 18tok/sQ5_K_M | 24tok/sQ4_K_M |
| Qwen 3.5 4B | 62tok/sQ8_0 | 79tok/sQ8_0 |
| Ministral 3 8B | 23tok/sQ8_0 | 29tok/sQ8_0 |
| Granite 4.2 8B | 21tok/sQ8_0 | 26tok/sQ8_0 |
| Gemma 4 E4B | 41tok/sQ8_0 | 52tok/sQ8_0 |
| LFM2.5 8B-A1B | 58tok/sQ8_0 | 74tok/sQ8_0 |
| Llama 3.1 8B | 27tok/sQ6_K | 35tok/sQ6_K |
| Qwen2.5 0.5B | 312tok/sQ8_0 | 395tok/sQ8_0 |
| Qwen3.5 0.8B | 255tok/sQ8_0 | 323tok/sQ8_0 |
| Qwen3.5 2B | 184tok/sQ5_K_M | 234tok/sQ5_K_M |
| DeepSeek R1 Distill Qwen 1.5B | 110tok/sQ8_0 | 140tok/sQ8_0 |
| MiniCPM5 2B | 76tok/sQ8_0 | 96tok/sQ8_0 |
| Spark X2.5 4B | 76tok/smlx-4bit | 96tok/smlx-4bit |
| Mellum2.1 12B A2.5B Thinking | 69tok/sQ8_0 | 87tok/sQ8_0 |
| FrogNano 4B 2609 | 61tok/sQ8_0 | 77tok/sQ8_0 |
| LFM2.5 2.6B | 60tok/sQ8_0 | 77tok/sQ8_0 |
| LFM2.5 VL 3B | 60tok/sQ8_0 | 77tok/sQ8_0 |
| Kimi VL A3B Thinking 2506 | 49tok/sQ8_0 | 62tok/sQ8_0 |
| Kimi VL A3B | 47tok/smlx-8bit | 60tok/smlx-8bit |
| granite vision 4.1 4b | 46tok/sQ8_0 | 59tok/sQ8_0 |
| granite 4.2 3b | 45tok/sQ8_0 | 57tok/sQ8_0 |
| Ministral 3 3B Instruct 2512 | 39tok/sQ8_0 | 50tok/sQ8_0 |
| GLM 4 9B 0414 | 38tok/smlx-5bit | 48tok/smlx-5bit |
| LFM2.5 24B A2B | 36tok/sQ5_K_M | 48tok/sQ4_K_M |
| DeepSeek R1 Distill Qwen 7B | 35tok/sQ8_0 | 45tok/sQ8_0 |
| GLM 4.6V Flash | 33tok/sQ8_0 | 42tok/sQ8_0 |
| GLM 4.1V 9B Thinking | 29tok/smlx-8bit | 37tok/smlx-8bit |
| DeepSeek R1 Distill Llama 8B | 24tok/sQ8_0 | 31tok/sQ8_0 |
| Ministral 3 8B Reasoning 2512 | 23tok/sQ8_0 | 29tok/sQ8_0 |
| Muse Glimmer 30B | 20tok/sQ5_K_M | 25tok/sQ5_K_M |
| Ministral 3 14B Reasoning 2512 | 19tok/sQ6_K | 24tok/sQ6_K |
| DeepSeek R1 Distill Qwen 14B | 18tok/sQ5_K_M | 23tok/sQ5_K_M |
| Qwen 3.5 122B-A10B | offload | offload |
| Gemma 4 31B | offload | offload |
| Qwen3 Coder Next | offload | offload |
| Mistral Small 4 119B | offload | offload |
| Qwen3 Next 80B-A3B | offload | offload |
| Devstral Small 2 24B | offload | offload |
| Kimi Linear 48B-A3B | offload | offload |
| Qwen3 Coder 30B-A3B | offload | offload |
| Qwen3 32B | offload | offload |
| Mistral Small 3.2 24B | offload | offload |
| DeepSeek R1 Distill Qwen 32B | offload | offload |
| Devstral Small 2507 | offload | offload |
| Magistral Small 2506 | offload | offload |
| Magistral Small 2507 | offload | offload |
| Magistral Small 2509 | offload | offload |
| Qwen3.8 27B OBLITERATED | offload | offload |
| granite 4.2 30b | 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 |
| gpt-oss 120B | won't run | won't run |
| GLM 4.5 Air | won't run | won't run |
| Llama 3.3 70B | won't run | won't run |
| DeepSeek R1 | won't run | won't run |
| DeepSeek R1 0528 | won't run | won't run |
| DeepSeek R1 Distill Llama 70B | won't run | won't run |
| DeepSeek R1 Zero | won't run | won't run |
| DeepSeek V3 | won't run | won't run |
| DeepSeek V3 0324 | won't run | won't run |
| DeepSeek V3.1 | won't run | won't run |
| DeepSeek V3.1 Terminus | won't run | won't run |
| DeepSeek V3.2 | won't run | won't run |
| DeepSeek V4 Flash 0731 | won't run | won't run |
| DeepSeek V4 Flash Vision Exp | won't run | won't run |
| Devstral 2 123B Instruct 2512 | won't run | won't run |
| GLM 4.5 | won't run | won't run |
| GLM 4.5V | won't run | won't run |
| GLM 4.6 | won't run | won't run |
| GLM 4.6V | won't run | won't run |
| GLM 4.7 | won't run | won't run |
| GLM 5 | won't run | won't run |
| GLM 5.1 | won't run | won't run |
| GLM 5.2 | won't run | won't run |
| GLM 5.3 | won't run | won't run |
| Kimi Dev 72B | won't run | won't run |
| Kimi K2 | won't run | won't run |
| Kimi K2 Instruct 0905 | won't run | won't run |
| Kimi K2 Thinking | won't run | won't run |
| Kimi K2.5 | won't run | won't run |
| MiniMax M2 | won't run | won't run |
| MiniMax M2.1 | won't run | won't run |
| MiniMax M2.5 | won't run | won't run |
| MiniMax M3 | won't run | won't run |
| Mistral Medium 3.5 128B | won't run | won't run |
Questions
- Which is better for local AI, the NVIDIA RTX PRO 4000 Blackwell or the AMD Radeon RX 7900 XTX?
- Choose the NVIDIA RTX PRO 4000 Blackwell to run bigger models. Choose the AMD Radeon RX 7900 XTX for speed.
- Which can run bigger models, the NVIDIA RTX PRO 4000 Blackwell or the AMD Radeon RX 7900 XTX?
- At 64k context the NVIDIA RTX PRO 4000 Blackwell runs 48 of the models we track fully in memory and the AMD Radeon RX 7900 XTX 47; only the NVIDIA RTX PRO 4000 Blackwell holds Nemotron 3.5 Lightning 30B-A3B.
- Which is faster for local AI, the NVIDIA RTX PRO 4000 Blackwell or the AMD Radeon RX 7900 XTX?
- On the models both run fully in memory, the AMD Radeon RX 7900 XTX is about 1.3× as fast as the NVIDIA RTX PRO 4000 Blackwell (estimated, median across models).
More comparisons
- Intel Arc Pro B60 vs NVIDIA RTX PRO 4000 Blackwell
- Intel Arc Pro B70 vs NVIDIA RTX PRO 4000 Blackwell
- AMD Radeon AI PRO R9700 vs NVIDIA RTX PRO 4000 Blackwell
- AMD Radeon AI PRO R9700 vs AMD Radeon RX 7900 XTX
- NVIDIA GeForce RTX 3090 vs NVIDIA RTX PRO 4000 Blackwell
- NVIDIA GeForce RTX 3090 vs AMD Radeon RX 7900 XTX
- NVIDIA GeForce RTX 3090 Ti vs AMD Radeon RX 7900 XTX
- NVIDIA GeForce RTX 4070 Ti Super vs NVIDIA RTX PRO 4000 Blackwell
- NVIDIA GeForce RTX 4070 Ti Super vs AMD Radeon RX 7900 XTX
- NVIDIA GeForce RTX 4080 Super vs NVIDIA RTX PRO 4000 Blackwell