localLLMriglocal, on your table

model

NVIDIA Nemotron 3.5 Super VL: how much memory, and what runs it

Hybrid Mamba-2/Transformer: 88 blocks (40 mamba, 40 moe, 8 attention). The KV formula uses only those 8 attention blocks. It does not model the Mamba SSM state or the vision mmproj (+0.97–1.73 GB). 120B total and 12B active are the card's round billions, so paramsEstimated stays true. Local on Ollama as nemotron-3.5-super (default tag 80–90 GB). Speed is not estimated: the formula streams every weight and understates sparse MoE.

NVIDIA Nemotron 3.5 Super VL needs about 72.7 GB at Q4_K_M with 8K context (estimate). See which GPUs and Macs fit it, and EU prices. Speed is not estimated for this mixture-of-experts model.

QuantisationEstimated memory
Q472.7 GBestimate
Q586.1 GBestimate
Q8129.3 GBestimate
FP16240.9 GBestimate

At 8,192 tokens of context. Catalog quant rates are a formula. sourceUrl on that memory stays null.

Parameter count source: huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16. 120B total, 12B active. Memory math uses the total.

VRAM per quant and the cheapest rig for NVIDIA Nemotron 3.5 Super VL

Open in the calculator

Three setups

NVIDIA Nemotron 3.5 Super VL · Q4 · 8,192 ctx

Estimated memory

72.7 GBestimate

71.8 GB weights · 0.1 GB KV cache · 0.8 GB overhead

  1. 01 · Fastest

    Mac Studio M5 Ultra 96GB

    High bandwidth and enough memory for 70B Q4/Q5. FP16 70B does not fit.

    $7,444€6,649

    price as of 16 Oct 2026

    Belgian Apple Store

    Belgian Apple Store price for this config, from the 16 Oct 2026 fact-check.

    speed

    not estimated for MoE yet

    speed: not estimated for MoE yet

    needed

    72.7 GB

    full fit

    usable

    84.0 GB

    unified

    Why this one

    • The model fits entirely in VRAM or unified memory. No CPU offload.
    • Unified memory: weights and KV cache share one pool.
    • High memory bandwidth, so a higher tokens-per-second estimate.
    • One box, little noise, no discrete GPU.
    • Highest estimated speed in this shortlist.

    Parts

    • Mac Studio M5 Ultra, 96 GB, 1.2 TB/s

  2. 02 · A step up

    Mac Studio M5 Ultra 256GB

    The machine in this list that can still hold 70B in FP16. The euro amount is an estimate.

    $13,994€12,500

    price as of 16 Oct 2026

    price unverified

    Estimate from the 16 Oct 2026 fact-check. Not a checked Apple Store listing. About €12,500–€13,000. This row stores the low end. 256 GB requires the 36-core CPU / 80-core GPU chip.

    speed

    not estimated for MoE yet

    speed: not estimated for MoE yet

    needed

    72.7 GB

    full fit

    usable

    244.0 GB

    unified

    Why this one

    • The model fits entirely in VRAM or unified memory. No CPU offload.
    • Memory left for a longer context or a heavier quantisation.
    • Unified memory: weights and KV cache share one pool.
    • High memory bandwidth, so a higher tokens-per-second estimate.
    • One box, little noise, no discrete GPU.

    Parts

    • Mac Studio M5 Ultra, 256 GB unified memory, 1.2 TB/s
    • 256 GB requires the 36-core CPU / 80-core GPU chip

  3. 03 · A step up

    Mac Studio M5 Max 128GB

    70B up to Q8 has room here. Bandwidth still trails a dual high-end GPU.

    $6,717€6,000

    price as of 16 Oct 2026

    price unverified

    Estimate from the 16 Oct 2026 fact-check. Not a checked Apple Store listing. About €6,000.

    speed

    not estimated for MoE yet

    speed: not estimated for MoE yet

    needed

    72.7 GB

    full fit

    usable

    116.0 GB

    unified

    Why this one

    • The model fits entirely in VRAM or unified memory. No CPU offload.
    • Memory left for a longer context or a heavier quantisation.
    • Unified memory: weights and KV cache share one pool.
    • One box, little noise, no discrete GPU.

    Parts

    • Mac Studio M5 Max 18-core/40-core GPU, 128 GB, 614 GB/s

With CPU offload, and therefore slower, this could also work: Dual RTX 4090, 48 GB, speed: not estimated for MoE yet, $8,206€7,330.

Tokens per second are estimated from memory bandwidth, not from a measured benchmark. EUR prices include VAT. USD is an FX estimate. A price without a checked listing is marked price unverified. Run llama-bench yourself before you buy.

Want a build tailored to you? Custom report

Get new builds when GPU prices move

End of results

Embed this

[![NVIDIA Nemotron 3.5 Super VL Q4 on RTX 3060 12GB: ✗ · needs 72.7 GB (estimate)](https://localllmrig.com/badge/nemotron-3.5-super-vl/rtx-3060-12gb.svg)](https://localllmrig.com/can-i-run/nemotron-3.5-super-vl/rtx-3060-12gb?utm_source=badge)

Model badge

[![NVIDIA Nemotron 3.5 Super VL Q4: needs ~72.7 GB VRAM (estimate)](https://localllmrig.com/badge/nemotron-3.5-super-vl.svg)](https://localllmrig.com/models/nemotron-3.5-super-vl?utm_source=badge)

Common questions

How much memory does NVIDIA Nemotron 3.5 Super VL need at Q4?
About 72.7 GB at 8,192 tokens: 71.8 GB weights, 0.1 GB KV cache, 0.8 GB overhead. That is an estimate. sourceUrl stays null.
Does NVIDIA Nemotron 3.5 Super VL fit on a single 24 GB GPU?
Not in full. The estimate is 72.7 GB, and a 24 GB card leaves 23.4 GB usable. sourceUrl stays null.
Are the tokens per second measured?
speed: not estimated for MoE yet
NVIDIA Nemotron 3.5 Super VL: VRAM needed (Q4–FP16) & · localLLMrig