ProGen2-xlarge TPS calculator

Open weights Salesforce Research,Columbia University,Johns Hopkins University 6.4B parameters June 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

589 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Tesla K20c

5 GB · IQ4_XS · 28.7 tok/s

Fastest card

B200

529 tok/s · 180 GB

Which GPUs can run ProGen2-xlarge?

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

589 cards match

Calculating

					Needs	Quantisation	Fit
529 tok/s 318–847 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	7.6 GB	Q8_0	Comfortable
529 tok/s 318–847 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	7.6 GB	Q8_0	Comfortable
423 tok/s 254–676 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	7.6 GB	Q8_0	Comfortable
423 tok/s 254–676 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	7.6 GB	Q8_0	Comfortable
338 tok/s 203–541 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	7.6 GB	Q8_0	Comfortable
324 tok/s 194–518 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	7.6 GB	Q8_0	Comfortable
324 tok/s 194–518 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	7.6 GB	Q8_0	Comfortable
310 tok/s 186–496 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	7.6 GB	Q8_0	Comfortable
275 tok/s 165–440 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	7.6 GB	Q8_0	Comfortable
275 tok/s 165–440 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	7.6 GB	Q8_0	Comfortable
275 tok/s 165–440 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	7.6 GB	Q8_0	Comfortable
261 tok/s 156–417 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	7.6 GB	Q8_0	Comfortable
222 tok/s 133–356 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	7.6 GB	Q8_0	Comfortable
222 tok/s 133–356 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	7.6 GB	Q8_0	Comfortable
222 tok/s 133–356 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	7.6 GB	Q8_0	Comfortable
222 tok/s 133–356 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	7.6 GB	Q8_0	Comfortable
222 tok/s 133–356 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	7.6 GB	Q8_0	Comfortable
169 tok/s 102–271 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	7.6 GB	Q8_0	Comfortable
169 tok/s 102–271 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	7.6 GB	Q8_0	Comfortable
143 tok/s 86–229 · low confidence	CMP 170HX 8 GB NVIDIA	8 GB	1,490 GB/s	Sep 2021	6.1 GB	Q6_K	Tight
141 tok/s 85–226 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	7.6 GB	Q8_0	Comfortable
138 tok/s 83–221 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	7.6 GB	Q8_0	Comfortable
135 tok/s 81–216 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	7.6 GB	Q8_0	Comfortable
135 tok/s 81–216 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	7.6 GB	Q8_0	Comfortable
135 tok/s 81–216 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	7.6 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: Salesforce Research,Columbia University,Johns Hopkins University
Organisation type: Industry,Academia,Academia
Country: United States of America
Published: 27 June 2022
Authors: Erik Nijkamp, Jeffrey Ruffolo, Eli N. Weinstein, Nikhil Naik, Ali Madani

What it does

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

Domain: Biology
Task: Proteins, Protein generation, Protein or nucleotide language model (pLM/nLM)
Approach: Self-supervised learning

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: 6.4B
Training data: 350,000,000,000 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: 1.4 × 10²² FLOP
How it was established: Hardware,Third-party estimation

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
Compute cost: $11,850

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: Confident
Citations: 507

Sources

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

Reference: ProGen2: Exploring the Boundaries of Protein Language Models
Last updated: 25 May 2026

The extremes

The ten fastest GPUs that run ProGen2-xlarge

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 ProGen2-xlarge

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 K20c

Memory needed

4.2 GB

Fastest

529 tok/s

ProGen2-xlarge is small enough at 6.4B parameters that hardware is rarely the obstacle — 589 of the cards we track can run it, including cards several years old.

At the low end, a Tesla K20c handles it — 5 GB, at IQ4_XS, for about 28.7 tokens per second.

The quickest result comes from a B200 at around 529 tokens per second — its 8,000 GB/s of bandwidth is what buys that.

About this model

ProGen2-xlarge was published by Salesforce Research,Columbia University,Johns Hopkins University, in United States of America, in June 2022. It comes out of industry,Academia,Academia.

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

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.

How fast it runs, and why

Across every card that can run it, the middle of the range is about 26.4 tokens per second, and 562 of them clear the ten tokens per second that roughly matches reading speed.

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.

Training and provenance

The training run consumed about 1.4 × 10²² FLOP, on Google TPU v3. That figure describes the cost of creating it and has no bearing on how quickly it generates text.

The training set ran to roughly 350,000,000,000 tokens.

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

Step by step

How to choose a GPU for ProGen2-xlarge

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

Every card here has been checked against ProGen2-xlarge — around 4.2 GB at IQ4_XS. Capacity is the gate — a card either holds it or it does not.
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: at long context ProGen2-xlarge can slip off a card that handles short questions easily.
03

Set a quality floor

The quantisation column varies by card, because a bigger card holds a more accurate copy of ProGen2-xlarge — IQ4_XS on the smallest card that fits. Set a floor to hold the comparison at one level.
04

Sort by speed

Ranking by tokens per second for ProGen2-xlarge follows memory bandwidth, not core counts, which is why the B200 tops it at 529 tok/s.
05

Look at the headroom, not just the fit

A tight fit runs ProGen2-xlarge but leaves nothing spare for a longer conversation; comfortable has headroom. If you expect to grow the context, buy for comfortable.
06

Open the card you have settled on

Following a card through to its own page shows every other model it can hold, which is the question that follows once ProGen2-xlarge is settled.

Answers

ProGen2-xlarge — common questions

Can I run ProGen2-xlarge if it does not fit in my GPU?

Partly. Layers that do not fit sit in system memory and run at a fraction of the speed, so a mostly-offloaded ProGen2-xlarge is rarely worth using — the nearest miss we calculate is short by 1.0 GB. Every figure here assumes the whole model is on the card.

Would two GPUs run ProGen2-xlarge faster?

Two cards buy memory rather than speed. That matters for ProGen2-xlarge only if one card cannot hold it — 589 can, so a second adds little.

Why does the quantisation differ between cards for ProGen2-xlarge?

A larger card holds a more accurate copy. Across the cards that run ProGen2-xlarge, 4 compression levels are used; the floor control above pins it to one.

How accurate are these ProGen2-xlarge speed estimates?

These are estimates with real error bars. The fastest result here, 318–847 tok/s on the B200, could reasonably land anywhere in its published range depending on which runtime you use.

What GPU do I need to run ProGen2-xlarge?

The smallest card in our catalogue that holds ProGen2-xlarge is the Tesla K20c, with 5 GB of memory. It runs the model at IQ4_XS using about 4.2 GB, and produces roughly 28.7 tokens per second. 589 cards in total can run it.

How fast is ProGen2-xlarge on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 529 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 562 of the cards that can run ProGen2-xlarge clear that.

How much VRAM does ProGen2-xlarge need?

About 4.2 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.

Can I run ProGen2-xlarge on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q6_K, using about 6.1 GB and generating roughly 143 tokens per second — a tight fit.

Can I run ProGen2-xlarge on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 7.6 GB and generating roughly 60.4 tokens per second — a comfortable fit.

Can I run ProGen2-xlarge on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 7.6 GB and generating roughly 74.8 tokens per second — a comfortable fit.

Can I run ProGen2-xlarge on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 7.6 GB and generating roughly 88.7 tokens per second — a comfortable fit.

Is ProGen2-xlarge open source?

Its weights are published, so ProGen2-xlarge 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 ProGen2-xlarge have?

ProGen2-xlarge has 6.4B parameters. "We introduce a suite of protein language models, named ProGen2, that are scaled up to 6.4B 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.

Who created ProGen2-xlarge?

ProGen2-xlarge was published by Salesforce Research,Columbia University,Johns Hopkins University, based in United States of America, categorised as industry,Academia,Academia.

When was ProGen2-xlarge released?

ProGen2-xlarge was published in June 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.

What is ProGen2-xlarge used for?

ProGen2-xlarge works in Biology, and is recorded as handling proteins, Protein generation, Protein or nucleotide language model (pLM/nLM). Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Where can I download ProGen2-xlarge?

The weights for ProGen2-xlarge are published, though we do not hold a repository link for it. This site calculates hardware requirements rather than hosting model files.

How much compute was used to train ProGen2-xlarge?

Around 1.4 × 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.

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