MSA Transformer TPS calculator

Open weights Facebook AI Research,University of California (UC) Berkeley,New York University (NYU) 100M parameters February 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

818 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Tesla C1080

4 GB · Q8_0 · 369 tok/s

Fastest card

B200

33,882 tok/s · 180 GB

Which GPUs can run MSA Transformer?

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

818 cards match

Calculating

					Needs	Quantisation	Fit
33,882 tok/s 20,329–54,212 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	0.8 GB	Q8_0	Comfortable
33,882 tok/s 20,329–54,212 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	0.8 GB	Q8_0	Comfortable
27,056 tok/s 16,234–43,289 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	0.8 GB	Q8_0	Comfortable
27,056 tok/s 16,234–43,289 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	0.8 GB	Q8_0	Comfortable
21,638 tok/s 12,983–34,621 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	0.8 GB	Q8_0	Comfortable
20,711 tok/s 12,426–33,137 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	0.8 GB	Q8_0	Comfortable
20,711 tok/s 12,426–33,137 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	0.8 GB	Q8_0	Comfortable
19,821 tok/s 11,893–31,714 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	0.8 GB	Q8_0	Comfortable
17,591 tok/s 10,555–28,146 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	0.8 GB	Q8_0	Comfortable
17,591 tok/s 10,555–28,146 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	0.8 GB	Q8_0	Comfortable
17,591 tok/s 10,555–28,146 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	0.8 GB	Q8_0	Comfortable
16,687 tok/s 10,012–26,699 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	0.8 GB	Q8_0	Comfortable
14,231 tok/s 8,538–22,769 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	0.8 GB	Q8_0	Comfortable
14,231 tok/s 8,538–22,769 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	0.8 GB	Q8_0	Comfortable
14,231 tok/s 8,538–22,769 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	0.8 GB	Q8_0	Comfortable
14,231 tok/s 8,538–22,769 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	0.8 GB	Q8_0	Comfortable
14,231 tok/s 8,538–22,769 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	0.8 GB	Q8_0	Comfortable
10,836 tok/s 6,501–17,337 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	0.8 GB	Q8_0	Comfortable
10,836 tok/s 6,501–17,337 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	0.8 GB	Q8_0	Comfortable
9,030 tok/s 5,418–14,447 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	0.8 GB	Q8_0	Comfortable
8,837 tok/s 5,302–14,139 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	0.8 GB	Q8_0	Comfortable
8,640 tok/s 5,184–13,824 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	0.8 GB	Q8_0	Comfortable
8,640 tok/s 5,184–13,824 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	0.8 GB	Q8_0	Comfortable
8,640 tok/s 5,184–13,824 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	0.8 GB	Q8_0	Comfortable
8,640 tok/s 5,184–13,824 · low confidence	H100 CNX NVIDIA	80 GB	2,040 GB/s	Mar 2023	0.8 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 Research,University of California (UC) Berkeley,New York University (NYU)
Organisation type: Industry,Academia,Academia
Country: United States of America, France
Published: 13 February 2021
Authors: Roshan Rao, Jason Liu, Robert Verkuil, Joshua Meier, John F. Canny, Pieter Abbeel, Tom Sercu, Alexander Rives

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 contact and distance prediction, Protein folding prediction

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: 100M
Training data: 1,395,000,000,000 tokens

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: 5.5 × 10²¹ FLOP
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: NVIDIA Tesla V100 DGXS 32 GB
Chips used: 32
Power draw: 16.2 kW
Compute cost: $13,257

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

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
Record confidence: Likely
Citations: 644

Sources

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

Reference: MSA Transformer
Last updated: 1 January 2026

The extremes

The ten fastest GPUs that run MSA Transformer

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 MSA Transformer

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

The hardware side

Minimum card

Tesla C1080

Memory needed

0.8 GB

Fastest

33,882 tok/s

MSA Transformer reaches a parameter count of 100M. That is small enough that hardware is rarely the obstacle, including on cards several years old. The number of cards we track that can run it: 818.

At the low end it is handled by Tesla C1080, with a memory capacity of 4 GB, running it at a compression of Q8_0 and producing around 369 tokens per second.

The fastest we calculate for it is B200, generating roughly 33,882 tokens per second on the strength of a memory bandwidth of 8,000 GB/s.

Background

MSA Transformer was published by Facebook AI Research,University of California (UC) Berkeley,New York University (NYU), in the country recorded as United States of America, during February 2021. The category the publisher falls under is industry,Academia,Academia.

It works in the domain of Biology, and is recorded as performing the task of proteins, Protein or nucleotide language model (pLM/nLM), Protein contact and distance prediction, Protein folding prediction.

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

Reading the throughput figures

Half the cards that hold it manage more than 951.4 tokens per second. Clearing the ten tokens per second that roughly matches reading speed: 818 of them.

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.

How it was trained

The training run consumed about 5.5 × 10²¹ FLOP, on hardware recorded as NVIDIA Tesla V100 DGXS 32 GB. That figure measures what producing the model cost, and has no bearing on how fast it answers.

Training consumed a corpus of around 1,395,000,000,000 tokens of text.

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

Step by step

How to choose a GPU for MSA Transformer

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

01

Start from the memory column

Start from what it actually needs, which is the requirement of MSA Transformer, needing around 0.8 GB at a compression of Q8_0. No amount of processing power compensates for a card that cannot hold it.
02

Set the context length you will work at

The conversation occupies memory too, and grows as it goes. Set the slider to the length you expect, because at long context a card that handles short questions easily can be dropped by MSA Transformer.
03

Choose how far you will compress it

Compression is what makes a model fit smaller cards, at some cost in accuracy, reaching a compression of Q8_0 on the smallest card that fits. Setting a minimum quality drops the cards that only manage it by squeezing further than you would want, and holds the comparison at one level.
04

Compare tokens per second, not specifications

Sort by speed to see how cards rank for MSA Transformer. It will not match a gaming ordering, because generation is bound by memory bandwidth. The card topping the list is B200, at 33,882 tok/s.
05

Check the fit verdict before buying

A tight fit runs, but leaves nothing spare for a longer conversation, in the case of MSA Transformer. Comfortable means you can grow the context later. That difference matters more than a few tokens per second, so buy for comfortable if you expect to.
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. A card is usually bought for more than one model, so it is worth a look before buying for MSA Transformer.

Answers

MSA Transformer — common questions

MSA Transformer— can I run it on a GPU holding 24 GB?

Yes. The card GeForce RTX 5090 D V2, holding 24 GB, runs it at a compression of Q8_0, using about 0.8 GB and generating roughly 5,675 tokens per second. The fit is comfortable.

MSA Transformer— is it open source?

Its weights are published, so it 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.

MSA Transformer— how many parameters does it have?

It has a parameter count of 100M. "We train an MSA Transformer model with 100M 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.

MSA Transformer— who created it?

It was published by Facebook AI Research,University of California (UC) Berkeley,New York University (NYU), based in United States of America, an organisation categorised as industry,Academia,Academia.

MSA Transformer— when was it released?

It was published in February 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.

MSA Transformer— what is it used for?

It works in the domain of Biology, and is recorded as handling the task of proteins, Protein or nucleotide language model (pLM/nLM), Protein contact and distance prediction, Protein folding prediction. 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.

MSA Transformer— where can I download it?

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

MSA Transformer— how much compute was used to train it?

Training consumed around 5.5 × 10²¹ FLOP, on hardware recorded as NVIDIA Tesla V100 DGXS 32 GB. 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.

MSA Transformer— can I run it if it does not fit in my GPU?

It can be split between the card and system memory, but it generates painfully slowly that way. Every figure here assumes the whole model is resident on the card.

MSA Transformer— would two GPUs run it faster?

Capacity adds across cards; throughput does not. The number of cards already holding it on their own: 818. So a second card is rarely the answer here.

MSA Transformer— why does the quantisation differ between cards?

Each card is shown running the least-compressed copy it can hold. The number of distinct compression levels across the cards that fit it: 1. Bigger cards get the more accurate version, and the quality floor above pins the comparison to one level.

MSA Transformer— how accurate are these speed estimates?

These are estimates with real error bars, and any of them could reasonably land anywhere in its published range depending on which runtime you use. The fastest result here: 20,329–54,212 tok/s on B200. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

MSA Transformer— what GPU do I need to run it?

The smallest card in our catalogue that holds it is Tesla C1080, with a memory capacity of 4 GB. It runs the model at a compression of Q8_0 using about 0.8 GB, and produces roughly 369 tokens per second. The number of cards able to run it in total: 818.

MSA Transformer— how fast is it on a GPU?

It depends on the card. The quickest we calculate is B200, at about 33,882 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and the number of cards clearing that: 818.

MSA Transformer— how much VRAM does it need?

It needs about 0.8 GB at a compression of Q8_0, 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.

MSA Transformer— can I run it on a GPU holding 8 GB?

Yes. The card CMP 170HX 8 GB, holding 8 GB, runs it at a compression of Q8_0, using about 0.8 GB and generating roughly 6,311 tokens per second. The fit is comfortable.

MSA Transformer— can I run it on a GPU holding 12 GB?

Yes. The card GeForce RTX 3080 Ti, holding 12 GB, runs it at a compression of Q8_0, using about 0.8 GB and generating roughly 3,864 tokens per second. The fit is comfortable.

MSA Transformer— can I run it on a GPU holding 16 GB?

Yes. The card Tesla V100 SXM2 16 GB, holding 16 GB, runs it at a compression of Q8_0, using about 0.8 GB and generating roughly 4,786 tokens per second. The fit is comfortable.

Source

Original publication

Record last updated 1 January 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