Skip to content

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

QuantTotal GBVerdictWeight sizing
GGUF Q4_075.4ComfortableVerified file
GGUF Q8_0134.3ComfortableVerified file

Context capacity · GGUF Q8_0

ContextTotal GBVerdict
4K134.3Comfortable
16K134.4Comfortable
32K134.5Comfortable
128K135.3Comfortable

Generation speed

GGUF Q8_0: 99 tok/s · Bandwidth ceiling (not measured)

Other model / GPU pairs