Instinct MI250X 128GB vs M3 Pro 18GB
For running local LLMs · 1 report across 1 model
Tokens per second by model
| Model | Instinct MI250X 128GB | M3 Pro 18GB |
|---|---|---|
| Gemma 4up to 31B | — | 14.0n=1 |
For running local LLMs · 1 report across 1 model
| Model | Instinct MI250X 128GB | M3 Pro 18GB |
|---|---|---|
| Gemma 4up to 31B | — | 14.0n=1 |