NVIDIA Nemotron Cascade 2 30B-A3B
NVIDIA Nemotron Cascade 2 30B-A3B loads in 24 GB at Q6_K and decodes at the speed of its 3B active path, comfortably from 48 GB of memory — Macs included, starting at the MacBook Pro M5 Pro 48GB. Below is the memory math we ran for it, the cheapest card that fits, and the fastest machine we track.
How much memory does NVIDIA Nemotron Cascade 2 30B-A3B need?
NVIDIA Nemotron Cascade 2 30B-A3B is a 30B mixture-of-experts model (3B active per token) built for reasoning, math, agentic tasks. The weights alone take 24 GB at Q6_K; add context and the OS, and it belongs in the 48 GB memory class. Every number below comes from the ModelFit engine — speeds are estimates, labeled est., because we do not benchmark hardware ourselves.
KV = fp16 estimate (q8_0 cache roughly halves it). "Comfortable" = weights + KV within the engine's tiered budget (~70-85% of memory).
Which machines run NVIDIA Nemotron Cascade 2 30B-A3B?
Verdicts come from the same engine that powers our fit checker — memory capacity first, bandwidth for speed. How we compute these numbers · Hugging Face model card
More Nemotron models
Frequently asked questions
How much memory does NVIDIA Nemotron Cascade 2 30B-A3B need?
24 GB for the Q6_K weights, plus ~4.0 GB of KV-cache at 16k context — about 28.0 GB total. Comfortable from 48 GB of VRAM or unified memory.
Does NVIDIA Nemotron Cascade 2 30B-A3B run on a Mac?
Yes — from the MacBook Pro M5 Pro 48GB (~30 tok/s est.). Unified memory means the RAM budget is the only limit.
What is the cheapest GPU for NVIDIA Nemotron Cascade 2 30B-A3B?
The AMD Ryzen AI Max+ 395 (Strix Halo) is the cheapest tracked card that runs NVIDIA Nemotron Cascade 2 30B-A3B comfortably — ~19 tok/s est. at ~$3,847 used (as of 2026-07-31).
Cite this page
ModelFit: NVIDIA Nemotron Cascade 2 30B-A3B — specs, memory math and hardware verdicts. https://modelfit.io/models/nemotron-cascade-2-30b-a3b/ (dataset updated 2026-09-03, CC BY 4.0).