Apple M2 Ultra (76-core GPU, 192GB) × Mistral Nemo 12B
Can the Apple M2 Ultra (76-core GPU, 192GB) run Mistral Nemo 12B?
Yes
Yes. The Apple M2 Ultra (76-core GPU, 192GB) can run Mistral Nemo 12B at up to F16 (needs ~27.3 GB of 144 GB usable), at an estimated ~48.7 tokens/sec at Q4_K_M.
Speed is estimated from memory bandwidth calibrated with a measured 94.27 tok/s Llama 2 7B Q4_0 result (source). No direct public benchmark of this exact pair is recorded yet.
By quantization (8K context)
| Quant | Memory needed | Fits? | Est. tok/s |
|---|---|---|---|
| Q3_K_M | 8.3 GB | ✅ | ~60.6 |
| Q4_K_M | 9.8 GB | ✅ | ~48.7 |
| Q5_K_M | 11.1 GB | ✅ | ~41.5 |
| Q6_K | 12.5 GB | ✅ | ~35.8 |
| Q8_0 | 15.5 GB | ✅ | ~27.8 |
| F16 | 27.3 GB | ✅ | ~14.8 |
Fastest hardware for Mistral Nemo 12B
- NVIDIA GeForce RTX 5090 — ~150 tok/s at Q4_K_M
- NVIDIA GeForce RTX 4090 — ~96.2 tok/s at Q4_K_M
- AMD Radeon RX 7900 XTX — ~95.3 tok/s at Q4_K_M
- NVIDIA GeForce RTX 5080 — ~94.1 tok/s at Q4_K_M
- NVIDIA GeForce RTX 5070 Ti — ~91.4 tok/s at Q4_K_M
No suitable GPU? Rent one by the hour on Vast.ai or RunPod (affiliate links).
Get new models & benchmarks in your inbox
No spam. Unsubscribe anytime by replying to any email.