Nemotron-H 47B TPS calculator

Open weights NVIDIA 47B parameters April 2025

Each card below is assessed against this model at the context length and minimum quality you choose. Speed is an estimate for a single request, calculated from the card's memory bandwidth and the size of the model once compressed.

Calculated for this model

93 of 818 cards that can run it

Smallest card that fits

Radeon PRO V710

28 GB · Q3_K_M · 9.6 tok/s

Fastest card

B200

72.1 tok/s · 180 GB

Which GPUs can run Nemotron-H 47B?

Set the inputs, read the answer

Context length tokens

A longer conversation needs more memory, which can push this model off smaller cards.

Minimum quality

Hides cards that would only fit the model by compressing it below this point.

Search cards

Card memory

Also show cards that cannot run it

93 cards match

Calculating

					Needs	Quantisation	Fit
72.1 tok/s 43–115 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	51.0 GB	Q8_0	Comfortable
72.1 tok/s 43–115 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	51.0 GB	Q8_0	Comfortable
57.6 tok/s 35–92 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	51.0 GB	Q8_0	Comfortable
57.6 tok/s 35–92 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	51.0 GB	Q8_0	Comfortable
46.0 tok/s 28–74 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	51.0 GB	Q8_0	Comfortable
44.1 tok/s 26–71 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	51.0 GB	Q8_0	Comfortable
44.1 tok/s 26–71 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	51.0 GB	Q8_0	Comfortable
42.2 tok/s 25–67 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	51.0 GB	Q8_0	Comfortable
41.4 tok/s 25–66 · low confidence	DRIVE A100 PROD NVIDIA	32 GB	1,870 GB/s	May 2020	26.4 GB	IQ4_XS	Tight
41.4 tok/s 25–66 · low confidence	GRID A100A NVIDIA	32 GB	1,870 GB/s	May 2020	26.4 GB	IQ4_XS	Tight
39.6 tok/s 24–63 · low confidence	GeForce RTX 5090 NVIDIA	32 GB	1,790 GB/s	Jan 2025	26.4 GB	IQ4_XS	Tight
39.6 tok/s 24–63 · low confidence	GeForce RTX 5090 D NVIDIA	32 GB	1,790 GB/s	Jan 2025	26.4 GB	IQ4_XS	Tight
37.4 tok/s 22–60 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	51.0 GB	Q8_0	Comfortable
37.4 tok/s 22–60 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	51.0 GB	Q8_0	Comfortable
37.4 tok/s 22–60 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	51.0 GB	Q8_0	Comfortable
35.5 tok/s 21–57 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	51.0 GB	Q8_0	Comfortable
30.3 tok/s 18–48 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	51.0 GB	Q8_0	Comfortable
30.3 tok/s 18–48 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	51.0 GB	Q8_0	Comfortable
30.3 tok/s 18–48 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	51.0 GB	Q8_0	Comfortable
30.3 tok/s 18–48 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	51.0 GB	Q8_0	Comfortable
30.3 tok/s 18–48 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	51.0 GB	Q8_0	Comfortable
25.1 tok/s 15–40 · low confidence	A100 PCIe 40 GB NVIDIA	40 GB	1,560 GB/s	Jun 2020	34.6 GB	Q5_K_M	Tight
25.1 tok/s 15–40 · low confidence	A100 SXM4 40 GB NVIDIA	40 GB	1,560 GB/s	May 2020	34.6 GB	Q5_K_M	Tight
25.1 tok/s 15–40 · low confidence	A800 PCIe 40 GB NVIDIA	40 GB	1,560 GB/s	Nov 2022	34.6 GB	Q5_K_M	Tight
25.0 tok/s 15–40 · low confidence	Tesla PG500-216 NVIDIA	32 GB	1,130 GB/s	Nov 2019	26.4 GB	IQ4_XS	Tight

Speeds are estimates for a single request — one conversation at a time — calculated from memory bandwidth, model size and quantisation. Real throughput varies with the inference runtime and its version. Figures published by hardware vendors measure many simultaneous requests and are much higher.

On record

Full specification

Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.

Origin

Who built this model, where, and when it was published.

Organisation: NVIDIA
Organisation type: Industry
Country: United States of America
Published: 14 April 2025
Authors: NVIDIA: Aaron Blakeman, Aarti Basant, Abhinav Khattar, Adithya Renduchintala, Akhiad Bercovich, Aleksander Ficek, Alexis Bjorlin, Ali Taghibakhshi, Amala Sanjay Deshmukh, Ameya Sunil Mahabaleshwarkar, Andrew Tao, Anna Shors, Ashwath Aithal, Ashwin Poojary, Ayush Dattagupta, Balaram Buddharaju, Bobby Chen, Boris Ginsburg, Boxin Wang, Brandon Norick, Brian Butterfield, Bryan Catanzaro, Carlo del Mun…

What it does

The problem areas the model was built for. A model can carry several of each.

Domain: Language
Task: Language modeling/generation, Question answering, Translation, Quantitative reasoning, Code generation, Neural Architecture Search - NAS
Base model: Nemotron-H 56B
Numerical format: FP8

Size

How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.

Parameters: 47B
Training data: tokens
Epochs: 1

Training compute

The arithmetic performed to train the model, measured in floating-point operations. It is a measure of what the training run cost, not of how fast the finished model answers you.

How it was established: Operation counting
Fine-tuning compute: 1.8 × 10²² FLOP

Availability

Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.

Weights: Open — downloadable
Model access: Open weights (non-commercial)
Training code: Unreleased
Hugging Face: nvidia

How it is classified

Labels the source dataset applies when tracking notable models, and how confident it is in the entry.

Record confidence: Confident

Sources

Where this record came from and when it was last checked.

Reference: Nemotron-H: A Family of Accurate and Efficient Hybrid Mamba-Transformer Models
Last updated: 28 November 2025

The extremes

The ten fastest GPUs for Nemotron-H 47B

Ranked by estimated tokens per second, newest card first where speeds tie. Because generation is bound by memory bandwidth, this ordering follows bandwidth rather than any gaming benchmark.

The smallest GPUs that still run Nemotron-H 47B

The cheapest route in, by memory capacity. A tight fit runs the model but leaves nothing spare for a longer conversation.

What the numbers mean

What you need to run it

Minimum card

Radeon PRO V710

Memory needed

23.7 GB

Fastest

72.1 tok/s

With 47B parameters, Nemotron-H 47B lands in the range a serious desktop card can handle once the weights are compressed. 93 of the cards we track can run it.

The least hardware that works is a Radeon PRO V710. Its 28 GB is enough at Q3_K_M compression, giving roughly 9.6 tokens per second.

Top of the range is the B200, at roughly 72.1 tokens per second thanks to 8,000 GB/s of bandwidth.

About this model

Nemotron-H 47B was published by NVIDIA, in United States of America, in April 2025. It comes out of industry.

It works in Language, and is recorded as doing language modeling/generation, Question answering, Translation, Quantitative reasoning, Code generation, Neural Architecture Search - NAS.

Its starting point was Nemotron-H 56B — most models at this scale are adapted from an existing base rather than built from nothing.

The weights are published, so it can be downloaded and run on your own hardware indefinitely, offline, with no account attached. It is published under the nvidia organisation on Hugging Face.

How fast it runs, and why

The median result is around 18.2 tokens per second; 72 cards produce text faster than most people read it.

It is a dense model, so every parameter is read for every token produced. That makes speed track memory bandwidth almost exactly — a card with twice the bandwidth generates roughly twice as fast.

Memory here is estimated from size rather than computed from the architecture, which is not recorded for this model — the numbers are indicative rather than exact.

Step by step

How to choose a GPU for Nemotron-H 47B

The table above has already assessed every card we hold specifications for against this model. Getting to your answer takes six steps.

01

Read the memory figure first

Look at what Nemotron-H 47B actually needs — around 23.7 GB at Q3_K_M. No amount of processing power compensates for a card that cannot hold it.
02

Set the context length you will work at

Set the context to what you will actually use. The cache grows with the conversation, and it is the usual reason Nemotron-H 47B stops fitting a card that seemed fine.
03

Choose how far you will compress it

Compression is what makes Nemotron-H 47B fit smaller cards, at some cost in accuracy — Q3_K_M on the smallest card that fits. A minimum quality removes the ones that go too far.
04

Compare tokens per second, not specifications

The speed ordering for Nemotron-H 47B is effectively an ordering by memory bandwidth, which is why the B200 tops it at 72.1 tok/s.
05

Look at the headroom, not just the fit

A tight fit runs Nemotron-H 47B but leaves nothing spare for a longer conversation; comfortable has headroom. If you expect to grow the context, buy for comfortable.
06

See what else that card runs

Every card name links to its own page, which runs the same calculation across the whole model catalogue. Worth a look before buying for Nemotron-H 47B alone — a card is usually bought for more than one model.

Answers

Nemotron-H 47B — common questions

Would two GPUs run Nemotron-H 47B faster?

Two cards buy memory rather than speed. That matters for Nemotron-H 47B only if one card cannot hold it — 93 can, so a second adds little.

Why does the quantisation differ between cards for Nemotron-H 47B?

Because capacity varies, so does how hard Nemotron-H 47B has to be squeezed — 5 distinct levels appear in the table above. Set a minimum quality to compare at one.

How accurate are these Nemotron-H 47B speed estimates?

They are calculated from specifications rather than measured, and each carries a range — 43–115 tok/s on the B200, for instance. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

What GPU do I need to run Nemotron-H 47B?

The smallest card in our catalogue that holds Nemotron-H 47B is the Radeon PRO V710, with 28 GB of memory. It runs the model at Q3_K_M using about 23.7 GB, and produces roughly 9.6 tokens per second. 93 cards in total can run it.

How fast is Nemotron-H 47B on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 72.1 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 72 of the cards that can run Nemotron-H 47B clear that.

How much VRAM does Nemotron-H 47B need?

About 23.7 GB at Q3_K_M compression, which is what the smallest card that runs it uses. Less compression needs more: the figures in the memory column above are recalculated for each card, because each one holds the least-compressed version it can.

Is Nemotron-H 47B open source?

Its weights are published, so Nemotron-H 47B can be downloaded and run on your own hardware. Note that open weights is not the same as open source in the full sense — it says nothing about the training data, the training code, or the commercial terms attached.

How many parameters does Nemotron-H 47B have?

Nemotron-H 47B has 47B parameters. "We use MiniPuzzle to distill Nemotron-H-56B-Base to Nemotron-H-47B-Base, using only 63 billion training tokens and FP8 training". That figure is the total, and it is what decides how much memory the model needs — roughly half a gigabyte per billion at the compression most people use.

Who created Nemotron-H 47B?

Nemotron-H 47B was published by NVIDIA, based in United States of America, categorised as industry.

When was Nemotron-H 47B released?

Nemotron-H 47B was published in April 2025.

What is Nemotron-H 47B used for?

Nemotron-H 47B works in Language, and is recorded as handling language modeling/generation, Question answering, Translation, Quantitative reasoning, Code generation, Neural Architecture Search - NAS. These are the areas it was designed around; they describe intent rather than a hard boundary.

Where can I download Nemotron-H 47B?

Its weights are published under the nvidia organisation on Hugging Face. We do not host model files — this site calculates what hardware is needed to run them.

Can I run Nemotron-H 47B if it does not fit in my GPU?

It can be split between the card and system memory, but Nemotron-H 47B generates painfully slowly that way — the nearest miss we calculate is short by 7.5 GB. Nothing on this page assumes offloading.

Source

Original publication

Record last updated 28 November 2025

The other direction

Looking at it from the other side?

This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.

All AI models