GPU comparison
AMD Ryzen AI Max+ 395 “Strix Halo” (128GB) vs NVIDIA RTX 5070 (12GB) for local LLMs
How the two stack up for running open-source LLMs locally — memory, bandwidth, price, and how many of the 64 tracked models each one can run.
| AMD Ryzen AI Max+ 395 “Strix Halo” (128GB) | NVIDIA RTX 5070 (12GB) | |
|---|---|---|
| Memory | 128.0 GB | 12.0 GB |
| Bandwidth | 256 GB/s | 672 GB/s |
| Price (approx) | $2,999 | $629 |
| LLMs it runs | 57 of 64 | 27 of 64 |
| Best model it runs | Qwen3.8 Flash Next (MoE) · 45–67 tok/s | Devstral Small 2 24B · 45–68 tok/s |
The AMD Ryzen AI Max+ 395 “Strix Halo” (128GB) runs 30 more of the tracked models (57 vs 27), thanks to its 128.0 GB of memory. The NVIDIA RTX 5070 (12GB) has more memory bandwidth (672 GB/s), so it generates tokens faster at the same model and quant.