ProtT5-XXL-BFD TPS calculator

Open weights Technical University of Munich,Med AI Technology,NVIDIA,Oak Ridge National Laboratory,Google,Seoul National University 11B parameters May 2021

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

509 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Xeon Phi 5110P

8 GB · IQ4_XS · 19.7 tok/s

Fastest card

B200

308 tok/s · 180 GB

Which GPUs can run ProtT5-XXL-BFD?

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.

509 cards match

Calculating
Needs Quantisation Fit
308 tok/s

185–493 · low confidence

B200 NVIDIA 180 GB 8,000 GB/s Jan 2024 12.5 GB Q8_0 Comfortable
308 tok/s

185–493 · low confidence

B300 NVIDIA 288 GB 8,000 GB/s Sep 2025 12.5 GB Q8_0 Comfortable
246 tok/s

148–394 · low confidence

Radeon Instinct MI350X AMD 288 GB 8,190 GB/s Jan 2025 12.5 GB Q8_0 Comfortable
246 tok/s

148–394 · low confidence

Radeon Instinct MI355X AMD 288 GB 8,190 GB/s Jan 2025 12.5 GB Q8_0 Comfortable
197 tok/s

118–315 · low confidence

Radeon Instinct MI300 AMD 128 GB 6,550 GB/s Jan 2023 12.5 GB Q8_0 Comfortable
188 tok/s

113–301 · low confidence

H200 NVL NVIDIA 141 GB 4,890 GB/s Nov 2024 12.5 GB Q8_0 Comfortable
188 tok/s

113–301 · low confidence

H200 SXM 141 GB NVIDIA 141 GB 4,890 GB/s Nov 2024 12.5 GB Q8_0 Comfortable
180 tok/s

108–288 · low confidence

Radeon Instinct MI325X AMD 256 GB 6,000 GB/s Oct 2024 12.5 GB Q8_0 Comfortable
160 tok/s

96–256 · low confidence

Radeon Instinct MI300A AMD 128 GB 5,325 GB/s Dec 2023 12.5 GB Q8_0 Comfortable
160 tok/s

96–256 · low confidence

Radeon Instinct MI300X AMD 192 GB 5,325 GB/s Dec 2023 12.5 GB Q8_0 Comfortable
160 tok/s

96–256 · low confidence

Radeon Instinct MI308X AMD 192 GB 5,325 GB/s Dec 2023 12.5 GB Q8_0 Comfortable
152 tok/s

91–243 · low confidence

H100 NVL 94 GB NVIDIA 94 GB 3,940 GB/s Mar 2023 12.5 GB Q8_0 Comfortable
141 tok/s

85–225 · low confidence

CMP 170HX 8 GB NVIDIA 8 GB 1,490 GB/s Sep 2021 6.7 GB IQ4_XS Tight
129 tok/s

78–207 · low confidence

H100 PCIe 96 GB NVIDIA 96 GB 3,360 GB/s Mar 2023 12.5 GB Q8_0 Comfortable
129 tok/s

78–207 · low confidence

H100 SXM5 80 GB NVIDIA 80 GB 3,360 GB/s Oct 2022 12.5 GB Q8_0 Comfortable
129 tok/s

78–207 · low confidence

H100 SXM5 94 GB NVIDIA 94 GB 3,360 GB/s Mar 2023 12.5 GB Q8_0 Comfortable
129 tok/s

78–207 · low confidence

H100 SXM5 96 GB NVIDIA 96 GB 3,360 GB/s Mar 2023 12.5 GB Q8_0 Comfortable
129 tok/s

78–207 · low confidence

H800 SXM5 NVIDIA 80 GB 3,360 GB/s Mar 2023 12.5 GB Q8_0 Comfortable
107 tok/s

64–172 · low confidence

CMP 170HX 10 GB NVIDIA 10 GB 1,560 GB/s Sep 2021 8.6 GB Q5_K_M Tight
98.5 tok/s

59–158 · low confidence

Radeon Instinct MI250 AMD 128 GB 3,280 GB/s Nov 2021 12.5 GB Q8_0 Comfortable
98.5 tok/s

59–158 · low confidence

Radeon Instinct MI250X AMD 128 GB 3,280 GB/s Nov 2021 12.5 GB Q8_0 Comfortable
82.1 tok/s

49–131 · low confidence

Data Center GPU Max 1550 Intel 128 GB 3,280 GB/s Jan 2023 12.5 GB Q8_0 Comfortable
80.3 tok/s

48–129 · low confidence

Data Center GPU Max Subsystem Intel 128 GB 3,210 GB/s Jan 2023 12.5 GB Q8_0 Comfortable
78.6 tok/s

47–126 · low confidence

A100 SXM4 80 GB NVIDIA 80 GB 2,040 GB/s Nov 2020 12.5 GB Q8_0 Comfortable
78.6 tok/s

47–126 · low confidence

A100X NVIDIA 80 GB 2,040 GB/s Jun 2021 12.5 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
Technical University of Munich,Med AI Technology,NVIDIA,Oak Ridge National Laboratory,Google,Seoul National University
Organisation type
Academia,Industry,Government,Industry,Academia
Country
Germany, China, United States of America, Korea (Republic of)
Published
4 May 2021
Authors
Ahmed Elnaggar, Michael Heinzinger, Christian Dallago, Ghalia Rehawi, Yu Wang, Llion Jones, Tom Gibbs, Tamas Feher, Christoph Angerer, Martin Steinegger, Debsindhu Bhowmik, Burkhard Rost

What it does

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

Domain
Biology
Task
Proteins, Protein or nucleotide language model (pLM/nLM), Protein representation learning
Approach
Self-supervised learning
Numerical format
FP32

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
11B

Table 2

Training data
393,000,000,000 tokens

Table 1: 2122M proteins, 393B amino acids, 572 GB

Epochs
5

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.

Training compute
3.7 × 10²² FLOP

FLOP = 11B*2*(920k*512*4096) + 11B*4*(920k*512*4096), 920k steps using seq length 512 batch size 4096,

How it was established
Operation counting

The training run

What it physically took to train: which chips, how many, for how long, and what that drew from the wall.

Training hardware
Google TPU v3
Chips used
512
Power draw
466.5 kW
Compute cost
$43,025

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
Unreleased

Licensed under the Academic Free License version 3.0 The ProtTrans project is a open source project supported by various partner companies and research institutions. We are committed to share all our pre-trained models and knowledge. https://github.com/agemagician/ProtTrans

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
ProtTrans:Towards Cracking the Language of Life's Code Through Self-Supervised Learning
Last updated
28 November 2025

The extremes

What the numbers mean

The hardware side

Minimum card

Xeon Phi 5110P

Memory needed

6.7 GB

Fastest

308 tok/s

ProtT5-XXL-BFD is small enough at 11B parameters that hardware is rarely the obstacle — 509 of the cards we track can run it, including cards several years old.

At the low end, a Xeon Phi 5110P handles it — 8 GB, at IQ4_XS, for about 19.7 tokens per second.

A B200 is the fastest we calculate for it: about 308 tokens per second, from 8,000 GB/s of memory bandwidth.

Where it came from

ProtT5-XXL-BFD was published by Technical University of Munich,Med AI Technology,NVIDIA,Oak Ridge National Laboratory,Google,Seoul National University, in Germany, in May 2021. It comes out of academia,Industry,Government,Industry,Academia.

It works in Biology, and is recorded as doing proteins, Protein or nucleotide language model (pLM/nLM), Protein representation learning.

The weights are published, so it can be downloaded and run on your own hardware indefinitely, offline, with no account attached.

Understanding the speeds

Half the cards that hold it manage more than 21.2 tokens per second, and 460 exceed reading speed outright.

Every weight participates in every token here, so bandwidth is the whole story: the ranking below is effectively a ranking of memory throughput.

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.

Training and provenance

The training run consumed about 3.7 × 10²² FLOP, on Google TPU v3. That figure describes the cost of creating it and has no bearing on how quickly it generates text.

Around 393,000,000,000 tokens went into training it.

Step by step

How to choose a GPU for ProtT5-XXL-BFD

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

  1. 01

    Check what it needs before anything else

    The table lists every card that can hold ProtT5-XXL-BFD — around 6.7 GB at IQ4_XS. That figure, not the card's headline performance, is what decides whether it runs.

  2. 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 ProtT5-XXL-BFD stops fitting a card that seemed fine.

  3. 03

    Choose how far you will compress it

    Each card runs the least-compressed copy it can hold — IQ4_XS on the smallest card that fits. Setting a floor drops the cards that only manage ProtT5-XXL-BFD by squeezing it further than you would want.

  4. 04

    Compare tokens per second, not specifications

    Sort by speed to see how cards rank for ProtT5-XXL-BFD. It will not match a gaming ordering — generation is bound by memory bandwidth, which is why the B200 tops it at 308 tok/s.

  5. 05

    Read the fit column last

    Tight means ProtT5-XXL-BFD loads and works, with no room to raise the context later. Comfortable means you can. The difference matters more than a few tokens per second.

  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 ProtT5-XXL-BFD alone — a card is usually bought for more than one model.

Answers

ProtT5-XXL-BFD — common questions

01

How many parameters does ProtT5-XXL-BFD have?

ProtT5-XXL-BFD has 11B parameters. Table 2. 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.

02

Who created ProtT5-XXL-BFD?

ProtT5-XXL-BFD was published by Technical University of Munich,Med AI Technology,NVIDIA,Oak Ridge National Laboratory,Google,Seoul National University, based in Germany, categorised as academia,Industry,Government,Industry,Academia.

03

When was ProtT5-XXL-BFD released?

ProtT5-XXL-BFD was published in May 2021. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

04

What is ProtT5-XXL-BFD used for?

ProtT5-XXL-BFD works in Biology, and is recorded as handling proteins, Protein or nucleotide language model (pLM/nLM), Protein representation learning. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

05

Where can I download ProtT5-XXL-BFD?

The weights for ProtT5-XXL-BFD are published, though we do not hold a repository link for it. This site calculates hardware requirements rather than hosting model files.

06

How much compute was used to train ProtT5-XXL-BFD?

Around 3.7 × 10²² FLOP, on Google TPU v3. That measures what producing the model cost and says nothing about how quickly it answers once trained — inference speed comes from memory bandwidth, not from the training budget.

07

Can I run ProtT5-XXL-BFD if it does not fit in my GPU?

It can be split between the card and system memory, but ProtT5-XXL-BFD generates painfully slowly that way — the nearest miss we calculate is short by 2.0 GB. Nothing on this page assumes offloading.

08

Would two GPUs run ProtT5-XXL-BFD faster?

Two cards buy memory rather than speed. That matters for ProtT5-XXL-BFD only if one card cannot hold it — 509 can, so a second adds little.

09

Why does the quantisation differ between cards for ProtT5-XXL-BFD?

A larger card holds a more accurate copy. Across the cards that run ProtT5-XXL-BFD, 4 compression levels are used; the floor control above pins it to one.

10

How accurate are these ProtT5-XXL-BFD speed estimates?

These are estimates with real error bars. The fastest result here, 185–493 tok/s on the B200, could reasonably land anywhere in its published range depending on which runtime you use.

11

What GPU do I need to run ProtT5-XXL-BFD?

The smallest card in our catalogue that holds ProtT5-XXL-BFD is the Xeon Phi 5110P, with 8 GB of memory. It runs the model at IQ4_XS using about 6.7 GB, and produces roughly 19.7 tokens per second. 509 cards in total can run it.

12

How fast is ProtT5-XXL-BFD on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 308 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 460 of the cards that can run ProtT5-XXL-BFD clear that.

13

How much VRAM does ProtT5-XXL-BFD need?

About 6.7 GB at IQ4_XS 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.

14

Can I run ProtT5-XXL-BFD on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at IQ4_XS, using about 6.7 GB and generating roughly 141 tokens per second — a tight fit.

15

Can I run ProtT5-XXL-BFD on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q6_K, using about 9.9 GB and generating roughly 51.0 tokens per second — a tight fit.

16

Can I run ProtT5-XXL-BFD on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 12.5 GB and generating roughly 43.5 tokens per second — a tight fit.

17

Can I run ProtT5-XXL-BFD on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 12.5 GB and generating roughly 51.6 tokens per second — a comfortable fit.

18

Is ProtT5-XXL-BFD open source?

Its weights are published, so ProtT5-XXL-BFD 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.

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.