ResNeXt-101 Billion-scale TPS calculator

Open weights Facebook AI 193M parameters May 2019

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

818 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Tesla C1080

4 GB · Q8_0 · 191 tok/s

Fastest card

B200

17,556 tok/s · 180 GB

Which GPUs can run ResNeXt-101 Billion-scale?

Set the inputs, read the answer

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

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

818 cards match

Calculating
Needs Quantisation Fit
17,556 tok/s

10,533–28,089 · low confidence

B200 NVIDIA 180 GB 8,000 GB/s Jan 2024 0.9 GB Q8_0 Comfortable
17,556 tok/s

10,533–28,089 · low confidence

B300 NVIDIA 288 GB 8,000 GB/s Sep 2025 0.9 GB Q8_0 Comfortable
14,019 tok/s

8,411–22,430 · low confidence

Radeon Instinct MI350X AMD 288 GB 8,190 GB/s Jan 2025 0.9 GB Q8_0 Comfortable
14,019 tok/s

8,411–22,430 · low confidence

Radeon Instinct MI355X AMD 288 GB 8,190 GB/s Jan 2025 0.9 GB Q8_0 Comfortable
11,211 tok/s

6,727–17,938 · low confidence

Radeon Instinct MI300 AMD 128 GB 6,550 GB/s Jan 2023 0.9 GB Q8_0 Comfortable
10,731 tok/s

6,439–17,169 · low confidence

H200 NVL NVIDIA 141 GB 4,890 GB/s Nov 2024 0.9 GB Q8_0 Comfortable
10,731 tok/s

6,439–17,169 · low confidence

H200 SXM 141 GB NVIDIA 141 GB 4,890 GB/s Nov 2024 0.9 GB Q8_0 Comfortable
10,270 tok/s

6,162–16,432 · low confidence

Radeon Instinct MI325X AMD 256 GB 6,000 GB/s Oct 2024 0.9 GB Q8_0 Comfortable
9,115 tok/s

5,469–14,583 · low confidence

Radeon Instinct MI300A AMD 128 GB 5,325 GB/s Dec 2023 0.9 GB Q8_0 Comfortable
9,115 tok/s

5,469–14,583 · low confidence

Radeon Instinct MI300X AMD 192 GB 5,325 GB/s Dec 2023 0.9 GB Q8_0 Comfortable
9,115 tok/s

5,469–14,583 · low confidence

Radeon Instinct MI308X AMD 192 GB 5,325 GB/s Dec 2023 0.9 GB Q8_0 Comfortable
8,646 tok/s

5,188–13,834 · low confidence

H100 NVL 94 GB NVIDIA 94 GB 3,940 GB/s Mar 2023 0.9 GB Q8_0 Comfortable
7,373 tok/s

4,424–11,797 · low confidence

H100 PCIe 96 GB NVIDIA 96 GB 3,360 GB/s Mar 2023 0.9 GB Q8_0 Comfortable
7,373 tok/s

4,424–11,797 · low confidence

H100 SXM5 80 GB NVIDIA 80 GB 3,360 GB/s Oct 2022 0.9 GB Q8_0 Comfortable
7,373 tok/s

4,424–11,797 · low confidence

H100 SXM5 94 GB NVIDIA 94 GB 3,360 GB/s Mar 2023 0.9 GB Q8_0 Comfortable
7,373 tok/s

4,424–11,797 · low confidence

H100 SXM5 96 GB NVIDIA 96 GB 3,360 GB/s Mar 2023 0.9 GB Q8_0 Comfortable
7,373 tok/s

4,424–11,797 · low confidence

H800 SXM5 NVIDIA 80 GB 3,360 GB/s Mar 2023 0.9 GB Q8_0 Comfortable
5,614 tok/s

3,369–8,983 · low confidence

Radeon Instinct MI250 AMD 128 GB 3,280 GB/s Nov 2021 0.9 GB Q8_0 Comfortable
5,614 tok/s

3,369–8,983 · low confidence

Radeon Instinct MI250X AMD 128 GB 3,280 GB/s Nov 2021 0.9 GB Q8_0 Comfortable
4,679 tok/s

2,807–7,486 · low confidence

Data Center GPU Max 1550 Intel 128 GB 3,280 GB/s Jan 2023 0.9 GB Q8_0 Comfortable
4,579 tok/s

2,747–7,326 · low confidence

Data Center GPU Max Subsystem Intel 128 GB 3,210 GB/s Jan 2023 0.9 GB Q8_0 Comfortable
4,477 tok/s

2,686–7,163 · low confidence

A100 SXM4 80 GB NVIDIA 80 GB 2,040 GB/s Nov 2020 0.9 GB Q8_0 Comfortable
4,477 tok/s

2,686–7,163 · low confidence

A100X NVIDIA 80 GB 2,040 GB/s Jun 2021 0.9 GB Q8_0 Comfortable
4,477 tok/s

2,686–7,163 · low confidence

A800 SXM4 80 GB NVIDIA 80 GB 2,040 GB/s Aug 2022 0.9 GB Q8_0 Comfortable
4,477 tok/s

2,686–7,163 · low confidence

H100 CNX NVIDIA 80 GB 2,040 GB/s Mar 2023 0.9 GB Q8_0 Comfortable

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
Facebook AI
Organisation type
Industry
Country
United States of America
Published
2 May 2019
Authors
I. Zeki Yalniz, Hervé Jégou, Kan Chen, Manohar Paluri, Dhruv Mahajan

What it does

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

Domain
Vision
Task
Image classification

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
193M
Training data
90,000,000 tokens

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

non-commercial for weights: https://github.com/facebookresearch/semi-supervised-ImageNet1K-models

How it is classified

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

Why it is tracked
SOTA improvement

"We demonstrate the performance of our method on popular classification benchmarks for both images and videos and significantly outperforms the state of the art."

Citations
490

Sources

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

Reference
Billion-scale semi-supervised learning for image classification
Last updated
25 May 2026

The extremes

The ten fastest GPUs that run ResNeXt-101 Billion-scale

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.

  1. 01 B300 288 GB · 8,000 GB/s · Q8_0 17,556 tok/s
  2. 02 B200 180 GB · 8,000 GB/s · Q8_0 17,556 tok/s
  3. 03 Radeon Instinct MI350X 288 GB · 8,190 GB/s · Q8_0 14,019 tok/s
  4. 04 Radeon Instinct MI355X 288 GB · 8,190 GB/s · Q8_0 14,019 tok/s
  5. 05 Radeon Instinct MI300 128 GB · 6,550 GB/s · Q8_0 11,211 tok/s
  6. 06 H200 NVL 141 GB · 4,890 GB/s · Q8_0 10,731 tok/s
  7. 07 H200 SXM 141 GB 141 GB · 4,890 GB/s · Q8_0 10,731 tok/s
  8. 08 Radeon Instinct MI325X 256 GB · 6,000 GB/s · Q8_0 10,270 tok/s
  9. 09 Radeon Instinct MI300A 128 GB · 5,325 GB/s · Q8_0 9,115 tok/s
  10. 10 Radeon Instinct MI300X 192 GB · 5,325 GB/s · Q8_0 9,115 tok/s

The smallest GPUs that still run ResNeXt-101 Billion-scale

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

  1. 01 GeForce RTX 4010 4 GB · needs 0.9 GB · Q8_0 · comfortable 211 tok/s
  2. 02 RTX A400 4 GB · needs 0.9 GB · Q8_0 · comfortable 211 tok/s
  3. 03 RTX 500 Mobile Ada Generation 4 GB · needs 0.9 GB · Q8_0 · comfortable 281 tok/s
  4. 04 GeForce RTX 3050 A Mobile 4 GB · needs 0.9 GB · Q8_0 · comfortable 421 tok/s
  5. 05 Jetson Orin Nano 4 GB 4 GB · needs 0.9 GB · Q8_0 · comfortable 74.9 tok/s
  6. 06 Radeon RX 6450M 4 GB · needs 0.9 GB · Q8_0 · comfortable 219 tok/s
  7. 07 Radeon RX 6550M 4 GB · needs 0.9 GB · Q8_0 · comfortable 246 tok/s
  8. 08 Radeon RX 6550S 4 GB · needs 0.9 GB · Q8_0 · comfortable 219 tok/s
  9. 09 Arc A310 4 GB · needs 0.9 GB · Q8_0 · comfortable 177 tok/s
  10. 10 Arc Pro A30M 4 GB · needs 0.9 GB · Q8_0 · comfortable 183 tok/s

What the numbers mean

Hardware requirements in practice

Minimum card

Tesla C1080

Memory needed

0.9 GB

Fastest

17,556 tok/s

ResNeXt-101 Billion-scale is small enough at 193M parameters that hardware is rarely the obstacle — 818 of the cards we track can run it, including cards several years old.

The entry point is the Tesla C1080: 4 GB of memory, Q8_0 compression, roughly 191 tokens per second.

Top of the range is the B200, at roughly 17,556 tokens per second thanks to 8,000 GB/s of bandwidth.

About this model

ResNeXt-101 Billion-scale was published by Facebook AI, in United States of America, in May 2019. The organisation is categorised as industry.

It works in Vision, and is recorded as doing image classification.

Published weights mean the model runs on your machine rather than someone else's, which is what makes the hardware question below answerable at all.

How fast it runs, and why

The median result is around 493.0 tokens per second; 818 cards produce text faster than most people read it.

Being dense, it reads all of itself per token, which is why the ordering by speed below follows the ordering by memory bandwidth so closely.

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.

Training and provenance

The training set ran to roughly 90,000,000 tokens.

The reason it appears in this catalogue at all is sOTA improvement.

Step by step

How to choose a GPU for ResNeXt-101 Billion-scale

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

  1. 01

    Start from the memory column

    Every card here has been checked against ResNeXt-101 Billion-scale — around 0.9 GB at Q8_0. Capacity is the gate — a card either holds it or it does not.

  2. 02

    Match the context to your actual use

    Set the context to what you will actually use. The cache grows with the conversation, and it is the usual reason ResNeXt-101 Billion-scale stops fitting a card that seemed fine.

  3. 03

    Choose how far you will compress it

    Compression is what makes ResNeXt-101 Billion-scale fit smaller cards, at some cost in accuracy — Q8_0 on the smallest card that fits. A minimum quality removes the ones that go too far.

  4. 04

    Compare tokens per second, not specifications

    Ranking by tokens per second for ResNeXt-101 Billion-scale follows memory bandwidth, not core counts, which is why the B200 tops it at 17,556 tok/s.

  5. 05

    Check the fit verdict before buying

    A tight fit runs ResNeXt-101 Billion-scale but leaves nothing spare for a longer conversation; comfortable has headroom. If you expect to grow the context, buy for comfortable.

  6. 06

    Open the card you have settled on

    Every card name links to its own page, which runs the same calculation across the whole model catalogue. Worth a look before buying for ResNeXt-101 Billion-scale alone — a card is usually bought for more than one model.

Answers

ResNeXt-101 Billion-scale — common questions

01

What is ResNeXt-101 Billion-scale used for?

ResNeXt-101 Billion-scale works in Vision, and is recorded as handling image classification. A model can carry several of each, so these are the areas it was built for rather than a limit on what it will attempt.

02

Where can I download ResNeXt-101 Billion-scale?

The weights for ResNeXt-101 Billion-scale are published, though we do not hold a repository link for it. This site calculates hardware requirements rather than hosting model files.

03

Can I run ResNeXt-101 Billion-scale if it does not fit in my GPU?

Only by offloading, which is usually a false economy: the part in system memory drags the whole thing down. Our figures for ResNeXt-101 Billion-scale assume it is fully resident.

04

Would two GPUs run ResNeXt-101 Billion-scale faster?

A second card roughly doubles the memory available but not the generation rate. With 818 cards already able to run ResNeXt-101 Billion-scale alone, the case for pairing is weak.

05

Why does the quantisation differ between cards for ResNeXt-101 Billion-scale?

Because capacity varies, so does how hard ResNeXt-101 Billion-scale has to be squeezed — 1 distinct levels appear in the table above. Set a minimum quality to compare at one.

06

How accurate are these ResNeXt-101 Billion-scale speed estimates?

They are calculated from specifications rather than measured, and each carries a range — 10,533–28,089 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.

07

What GPU do I need to run ResNeXt-101 Billion-scale?

The smallest card in our catalogue that holds ResNeXt-101 Billion-scale is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 0.9 GB, and produces roughly 191 tokens per second. 818 cards in total can run it.

08

How fast is ResNeXt-101 Billion-scale on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 17,556 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 818 of the cards that can run ResNeXt-101 Billion-scale clear that.

09

How much VRAM does ResNeXt-101 Billion-scale need?

About 0.9 GB at Q8_0 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.

10

Can I run ResNeXt-101 Billion-scale on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q8_0, using about 0.9 GB and generating roughly 3,270 tokens per second — a comfortable fit.

11

Can I run ResNeXt-101 Billion-scale on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 0.9 GB and generating roughly 2,002 tokens per second — a comfortable fit.

12

Can I run ResNeXt-101 Billion-scale on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 0.9 GB and generating roughly 2,480 tokens per second — a comfortable fit.

13

Can I run ResNeXt-101 Billion-scale on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 0.9 GB and generating roughly 2,941 tokens per second — a comfortable fit.

14

Is ResNeXt-101 Billion-scale open source?

Its weights are published, so ResNeXt-101 Billion-scale 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.

15

How many parameters does ResNeXt-101 Billion-scale have?

ResNeXt-101 Billion-scale has 193M parameters. 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.

16

Who created ResNeXt-101 Billion-scale?

ResNeXt-101 Billion-scale was published by Facebook AI, based in United States of America, categorised as industry.

17

When was ResNeXt-101 Billion-scale released?

ResNeXt-101 Billion-scale was published in May 2019. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

Source

Original publication

Record last updated 25 May 2026

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