Hardware comparison
Mac Studio M5 Ultra (96GB) vs NVIDIA RTX Pro 6000 Blackwell (96GB) for local LLMs
How the two stack up for running open-source LLMs locally — memory, bandwidth, price, and how many of the 69 tracked models each one can run.
Scroll sideways to compare both machines. Keyboard: focus the table, then use the arrow keys. Metric names stay visible.
| Hardware | Mac Studio M5 Ultra (96GB) | NVIDIA RTX Pro 6000 Blackwell (96GB) |
|---|---|---|
| Memory | 96.0 GB | 96.0 GB |
| Bandwidth | 1228 GB/s | 1792 GB/s (better for this comparison) |
| Price (approx) | $5,499Complete system | $12,999GPU only · host PC required, sold separately |
| LLMs it runs | 60 of 69 | 62 of 69 (better for this comparison) |
| Best model it runs | Gemma 4 31BQ8_0 · estimated 21–32 tok/s | Gemma 4 31BFP16 · estimated 20–30 tok/s |
These prices cover different purchases: a GPU card excludes its required host PC; a complete system includes the machine. No price winner is marked. Prices are approximate catalog estimates; check current configurations and prices before buying.
The recommended quants differ. These estimated speeds are not a like-for-like comparison. Estimates use 4,096 context tokens, llama.cpp and one stream.
The NVIDIA RTX Pro 6000 Blackwell (96GB) runs 2 more of the tracked models (62 vs 60). With the same nominal memory, more of it is usable for model weights on this architecture. The NVIDIA RTX Pro 6000 Blackwell (96GB) has more memory bandwidth (1792 GB/s). Higher bandwidth can help generation speed, but model, quant, architecture and runtime also affect the estimates.