Can Mac M5 Ultra 512G run Nemotron 3.5 Super VL 120B-A12B?
Comfortable: GGUF Q8_0 needs 134.3 GB at 4K, batch 1; headroom 249.7 GB.
Text input, FP16 KV cache, one card. Unified-memory capacity uses the GPU guide’s usable limit; negative headroom may require host offload. Speed source settings may differ from this 4K estimate.
Measured runs
No measured run for this pair. Sizing is an estimate, not a runtime test.
Quant memory at 4K
| Quant | Total GB | Verdict | Weight sizing |
|---|---|---|---|
| GGUF Q4_0 | 75.4 | Comfortable | Verified file |
| GGUF Q8_0 | 134.3 | Comfortable | Verified file |
Context capacity · GGUF Q8_0
| Context | Total GB | Verdict |
|---|---|---|
| 4K | 134.3 | Comfortable |
| 16K | 134.4 | Comfortable |
| 32K | 134.5 | Comfortable |
| 128K | 135.3 | Comfortable |
Generation speed
GGUF Q8_0: 99 tok/s · Bandwidth ceiling (not measured)