data2vec (language) TPS calculator

Open weights Meta AI 705.1M parameters January 2022

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 · 52.3 tok/s

Fastest card

B200

4,805 tok/s · 180 GB

Which GPUs can run data2vec (language)?

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
4,805 tok/s

2,883–7,688 · low confidence

B200 NVIDIA 180 GB 8,000 GB/s Jan 2024 1.5 GB Q8_0 Comfortable
4,805 tok/s

2,883–7,688 · low confidence

B300 NVIDIA 288 GB 8,000 GB/s Sep 2025 1.5 GB Q8_0 Comfortable
3,837 tok/s

2,302–6,139 · low confidence

Radeon Instinct MI350X AMD 288 GB 8,190 GB/s Jan 2025 1.5 GB Q8_0 Comfortable
3,837 tok/s

2,302–6,139 · low confidence

Radeon Instinct MI355X AMD 288 GB 8,190 GB/s Jan 2025 1.5 GB Q8_0 Comfortable
3,069 tok/s

1,841–4,910 · low confidence

Radeon Instinct MI300 AMD 128 GB 6,550 GB/s Jan 2023 1.5 GB Q8_0 Comfortable
2,937 tok/s

1,762–4,699 · low confidence

H200 NVL NVIDIA 141 GB 4,890 GB/s Nov 2024 1.5 GB Q8_0 Comfortable
2,937 tok/s

1,762–4,699 · low confidence

H200 SXM 141 GB NVIDIA 141 GB 4,890 GB/s Nov 2024 1.5 GB Q8_0 Comfortable
2,811 tok/s

1,687–4,498 · low confidence

Radeon Instinct MI325X AMD 256 GB 6,000 GB/s Oct 2024 1.5 GB Q8_0 Comfortable
2,495 tok/s

1,497–3,992 · low confidence

Radeon Instinct MI300A AMD 128 GB 5,325 GB/s Dec 2023 1.5 GB Q8_0 Comfortable
2,495 tok/s

1,497–3,992 · low confidence

Radeon Instinct MI300X AMD 192 GB 5,325 GB/s Dec 2023 1.5 GB Q8_0 Comfortable
2,495 tok/s

1,497–3,992 · low confidence

Radeon Instinct MI308X AMD 192 GB 5,325 GB/s Dec 2023 1.5 GB Q8_0 Comfortable
2,367 tok/s

1,420–3,786 · low confidence

H100 NVL 94 GB NVIDIA 94 GB 3,940 GB/s Mar 2023 1.5 GB Q8_0 Comfortable
2,018 tok/s

1,211–3,229 · low confidence

H100 PCIe 96 GB NVIDIA 96 GB 3,360 GB/s Mar 2023 1.5 GB Q8_0 Comfortable
2,018 tok/s

1,211–3,229 · low confidence

H100 SXM5 80 GB NVIDIA 80 GB 3,360 GB/s Oct 2022 1.5 GB Q8_0 Comfortable
2,018 tok/s

1,211–3,229 · low confidence

H100 SXM5 94 GB NVIDIA 94 GB 3,360 GB/s Mar 2023 1.5 GB Q8_0 Comfortable
2,018 tok/s

1,211–3,229 · low confidence

H100 SXM5 96 GB NVIDIA 96 GB 3,360 GB/s Mar 2023 1.5 GB Q8_0 Comfortable
2,018 tok/s

1,211–3,229 · low confidence

H800 SXM5 NVIDIA 80 GB 3,360 GB/s Mar 2023 1.5 GB Q8_0 Comfortable
1,537 tok/s

922–2,459 · low confidence

Radeon Instinct MI250 AMD 128 GB 3,280 GB/s Nov 2021 1.5 GB Q8_0 Comfortable
1,537 tok/s

922–2,459 · low confidence

Radeon Instinct MI250X AMD 128 GB 3,280 GB/s Nov 2021 1.5 GB Q8_0 Comfortable
1,281 tok/s

768–2,049 · low confidence

Data Center GPU Max 1550 Intel 128 GB 3,280 GB/s Jan 2023 1.5 GB Q8_0 Comfortable
1,253 tok/s

752–2,005 · low confidence

Data Center GPU Max Subsystem Intel 128 GB 3,210 GB/s Jan 2023 1.5 GB Q8_0 Comfortable
1,225 tok/s

735–1,960 · low confidence

A100 SXM4 80 GB NVIDIA 80 GB 2,040 GB/s Nov 2020 1.5 GB Q8_0 Comfortable
1,225 tok/s

735–1,960 · low confidence

A100X NVIDIA 80 GB 2,040 GB/s Jun 2021 1.5 GB Q8_0 Comfortable
1,225 tok/s

735–1,960 · low confidence

A800 SXM4 80 GB NVIDIA 80 GB 2,040 GB/s Aug 2022 1.5 GB Q8_0 Comfortable
1,225 tok/s

735–1,960 · low confidence

H100 CNX NVIDIA 80 GB 2,040 GB/s Mar 2023 1.5 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
Meta AI
Organisation type
Industry
Country
United States of America
Published
20 January 2022
Authors
Alexei Baevski, Wei-Ning Hsu, Qiantong Xu , Arun Babu, Jiatao Gu, Michael Auli

What it does

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

Domain
Language
Task
Language modeling/generation
Approach
Self-supervised learning
Numerical format
FP16

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
705.1M

Section 4: "We experiment with two model sizes: data2vec Base and data2vec Large, containing either L = 12 or L = 24 Trans- former blocks with H = 768 or H = 1024 hidden dimen- sion (with 4 × H feed-forward inner-dimension)"

Training data
131,072,000,000 tokens

Section 5.3: "we adopt the same training setup as BERT (Devlin et al., 2019) by pre-training on the Books Corpus (Zhu et al., 2015) and English Wikipedia data over 1M updates and a batch size of 256 sequences."

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 (unrestricted)
Training code
Open source

MIT License Models and code are available at www.github.com/pytorch/fairseq/tree/master/examples/data2vec

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

"Experiments on the major benchmarks of speech recognition, image classification, and natural lan guage understanding demonstrate a new state of the art or competitive performance to predominant approaches" "To our knowledge this is the first successful pre-trained NLP model which does not use discrete units (words, subwords, characters or bytes) as the training target. Instead, the model predicts a contextualized latent representation emerging from self-attention over the entire unmasked text …

Record confidence
Likely
Citations
1,021

Sources

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

Reference
Data2vec: A General Framework for Self-supervised Learning in Speech, Vision and Language
Last updated
11 February 2026

The extremes

The ten fastest GPUs that run data2vec (language)

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 4,805 tok/s
  2. 02 B200 180 GB · 8,000 GB/s · Q8_0 4,805 tok/s
  3. 03 Radeon Instinct MI350X 288 GB · 8,190 GB/s · Q8_0 3,837 tok/s
  4. 04 Radeon Instinct MI355X 288 GB · 8,190 GB/s · Q8_0 3,837 tok/s
  5. 05 Radeon Instinct MI300 128 GB · 6,550 GB/s · Q8_0 3,069 tok/s
  6. 06 H200 NVL 141 GB · 4,890 GB/s · Q8_0 2,937 tok/s
  7. 07 H200 SXM 141 GB 141 GB · 4,890 GB/s · Q8_0 2,937 tok/s
  8. 08 Radeon Instinct MI325X 256 GB · 6,000 GB/s · Q8_0 2,811 tok/s
  9. 09 Radeon Instinct MI300A 128 GB · 5,325 GB/s · Q8_0 2,495 tok/s
  10. 10 Radeon Instinct MI300X 192 GB · 5,325 GB/s · Q8_0 2,495 tok/s

The smallest GPUs that still run data2vec (language)

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 1.5 GB · Q8_0 · comfortable 57.7 tok/s
  2. 02 RTX A400 4 GB · needs 1.5 GB · Q8_0 · comfortable 57.7 tok/s
  3. 03 RTX 500 Mobile Ada Generation 4 GB · needs 1.5 GB · Q8_0 · comfortable 76.9 tok/s
  4. 04 GeForce RTX 3050 A Mobile 4 GB · needs 1.5 GB · Q8_0 · comfortable 115 tok/s
  5. 05 Jetson Orin Nano 4 GB 4 GB · needs 1.5 GB · Q8_0 · comfortable 20.5 tok/s
  6. 06 Radeon RX 6450M 4 GB · needs 1.5 GB · Q8_0 · comfortable 60.0 tok/s
  7. 07 Radeon RX 6550M 4 GB · needs 1.5 GB · Q8_0 · comfortable 67.5 tok/s
  8. 08 Radeon RX 6550S 4 GB · needs 1.5 GB · Q8_0 · comfortable 60.0 tok/s
  9. 09 Arc A310 4 GB · needs 1.5 GB · Q8_0 · comfortable 48.4 tok/s
  10. 10 Arc Pro A30M 4 GB · needs 1.5 GB · Q8_0 · comfortable 50.0 tok/s

What the numbers mean

What it takes to run this model

Minimum card

Tesla C1080

Memory needed

1.5 GB

Fastest

4,805 tok/s

data2vec (language) is small enough at 705.1M parameters that hardware is rarely the obstacle — 818 of the cards we track can run it, including cards several years old.

The smallest card that holds it is the Tesla C1080 with 4 GB, running it at Q8_0 and producing around 52.3 tokens per second.

A B200 is the fastest we calculate for it: about 4,805 tokens per second, from 8,000 GB/s of memory bandwidth.

Where it came from

data2vec (language) was published by Meta AI, in United States of America, in January 2022. industry is the category the publisher falls under.

It works in Language, and is recorded as doing language modeling/generation.

The weights being open is what puts this page in the calculator rather than only in the catalogue: it is a model you can actually hold.

Understanding the speeds

Half the cards that hold it manage more than 134.9 tokens per second, and 809 exceed reading speed outright.

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.

Without the attention layout on record, the memory column is an approximation. It is close enough to choose hardware by, and least reliable at long context.

What went into building it

Around 131,072,000,000 tokens went into training it.

Its inclusion criterion is sOTA improvement.

Step by step

How to choose a GPU for data2vec (language)

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

    Look at what data2vec (language) actually needs — around 1.5 GB at Q8_0. No amount of processing power compensates for a card that cannot hold it.

  2. 02

    Set the context length you will work at

    Longer conversations cost memory on top of what the weights need. Move the slider to your real working length before trusting any row for data2vec (language).

  3. 03

    Set a quality floor

    Each card runs the least-compressed copy it can hold — Q8_0 on the smallest card that fits. Setting a floor drops the cards that only manage data2vec (language) by squeezing it further than you would want.

  4. 04

    Rank by throughput rather than spec sheet

    Ranking by tokens per second for data2vec (language) follows memory bandwidth, not core counts, which is why the B200 tops it at 4,805 tok/s.

  5. 05

    Look at the headroom, not just the fit

    The fit column separates cards that just manage data2vec (language) from those with room to spare. Buy for the second if the context might grow.

  6. 06

    Check the card from the other side

    Every card name links to its own page, which runs the same calculation across the whole model catalogue. Worth a look before buying for data2vec (language) alone — a card is usually bought for more than one model.

Answers

data2vec (language) — common questions

01

Can I run data2vec (language) 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 data2vec (language) assume it is fully resident.

02

Would two GPUs run data2vec (language) faster?

Two cards buy memory rather than speed. That matters for data2vec (language) only if one card cannot hold it — 818 can, so a second adds little.

03

Why does the quantisation differ between cards for data2vec (language)?

Each card is shown running the least-compressed copy it can hold, and data2vec (language) appears at 1 different compression levels across the cards that fit it. Bigger cards get the more accurate version.

04

How accurate are these data2vec (language) speed estimates?

They are calculated from specifications rather than measured, and each carries a range — 2,883–7,688 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.

05

What GPU do I need to run data2vec (language)?

The smallest card in our catalogue that holds data2vec (language) is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 1.5 GB, and produces roughly 52.3 tokens per second. 818 cards in total can run it.

06

How fast is data2vec (language) on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 4,805 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 809 of the cards that can run data2vec (language) clear that.

07

How much VRAM does data2vec (language) need?

About 1.5 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.

08

Can I run data2vec (language) on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q8_0, using about 1.5 GB and generating roughly 895 tokens per second — a comfortable fit.

09

Can I run data2vec (language) on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 1.5 GB and generating roughly 548 tokens per second — a comfortable fit.

10

Can I run data2vec (language) on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 1.5 GB and generating roughly 679 tokens per second — a comfortable fit.

11

Can I run data2vec (language) on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 1.5 GB and generating roughly 805 tokens per second — a comfortable fit.

12

Is data2vec (language) open source?

Its weights are published, so data2vec (language) 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.

13

How many parameters does data2vec (language) have?

data2vec (language) has 705.1M parameters. Section 4: "We experiment with two model sizes: data2vec Base and data2vec Large, containing either L = 12 or L = 24 Trans- former blocks with H = 768 or H = 1024 hidden dimen- sion (with 4 × H feed-forward inner-dimension)". 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.

14

Who created data2vec (language)?

data2vec (language) was published by Meta AI, based in United States of America, categorised as industry.

15

When was data2vec (language) released?

data2vec (language) was published in January 2022. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

16

What is data2vec (language) used for?

data2vec (language) works in Language, and is recorded as handling language modeling/generation. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

17

Where can I download data2vec (language)?

The weights for data2vec (language) are published, though we do not hold a repository link for it. This site calculates hardware requirements rather than hosting model files.

Source

Original publication

Record last updated 11 February 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