Rita-XLarge TPS calculator

Open weights LightOn,Harvard University,University of Oxford 1.2B parameters July 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 · 30.7 tok/s

Fastest card

B200

2,824 tok/s · 180 GB

Which GPUs can run Rita-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

818 cards match

Calculating

					Needs	Quantisation	Fit
2,824 tok/s 1,694–4,518 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	2.0 GB	Q8_0	Comfortable
2,824 tok/s 1,694–4,518 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	2.0 GB	Q8_0	Comfortable
2,255 tok/s 1,353–3,607 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	2.0 GB	Q8_0	Comfortable
2,255 tok/s 1,353–3,607 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	2.0 GB	Q8_0	Comfortable
1,803 tok/s 1,082–2,885 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	2.0 GB	Q8_0	Comfortable
1,726 tok/s 1,036–2,761 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	2.0 GB	Q8_0	Comfortable
1,726 tok/s 1,036–2,761 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	2.0 GB	Q8_0	Comfortable
1,652 tok/s 991–2,643 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	2.0 GB	Q8_0	Comfortable
1,466 tok/s 880–2,346 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	2.0 GB	Q8_0	Comfortable
1,466 tok/s 880–2,346 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	2.0 GB	Q8_0	Comfortable
1,466 tok/s 880–2,346 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	2.0 GB	Q8_0	Comfortable
1,391 tok/s 834–2,225 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	2.0 GB	Q8_0	Comfortable
1,186 tok/s 712–1,897 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	2.0 GB	Q8_0	Comfortable
1,186 tok/s 712–1,897 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	2.0 GB	Q8_0	Comfortable
1,186 tok/s 712–1,897 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	2.0 GB	Q8_0	Comfortable
1,186 tok/s 712–1,897 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	2.0 GB	Q8_0	Comfortable
1,186 tok/s 712–1,897 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	2.0 GB	Q8_0	Comfortable
903 tok/s 542–1,445 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	2.0 GB	Q8_0	Comfortable
903 tok/s 542–1,445 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	2.0 GB	Q8_0	Comfortable
752 tok/s 451–1,204 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	2.0 GB	Q8_0	Comfortable
736 tok/s 442–1,178 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	2.0 GB	Q8_0	Comfortable
720 tok/s 432–1,152 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	2.0 GB	Q8_0	Comfortable
720 tok/s 432–1,152 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	2.0 GB	Q8_0	Comfortable
720 tok/s 432–1,152 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	2.0 GB	Q8_0	Comfortable
720 tok/s 432–1,152 · low confidence	H100 CNX NVIDIA	80 GB	2,040 GB/s	Mar 2023	2.0 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: LightOn,Harvard University,University of Oxford
Organisation type: Industry,Academia,Academia
Country: France, United States of America, United Kingdom of Great Britain and Northern Ireland
Published: 14 July 2022
Authors: Daniel Hesslow, Niccolo Zanichelli, Pascal Notin, Iacopo Poli, Debora Marks

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)

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: 1.2B
Training data: 150,000,000,000 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: 8.6 × 10²⁰ FLOP
How it was established: Reported

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 V100

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.

Record confidence: Confident
Citations: 128

Sources

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

Reference: RITA: a Study on Scaling Up Generative Protein Sequence Models
Last updated: 25 May 2026

The extremes

The ten fastest GPUs that run Rita-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 Rita-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

Hardware requirements in practice

Minimum card

Tesla C1080

Memory needed

2.0 GB

Fastest

2,824 tok/s

Rita-XLarge is small enough at 1.2B 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 30.7 tokens per second.

A B200 is the fastest we calculate for it: about 2,824 tokens per second, from 8,000 GB/s of memory bandwidth.

About this model

Rita-XLarge was published by LightOn,Harvard University,University of Oxford, in France, in July 2022. The organisation is categorised as industry,Academia,Academia.

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

Because its weights were released, nothing about running it depends on a provider staying available — it is yours once downloaded.

How fast it runs, and why

Half the cards that hold it manage more than 79.3 tokens per second, and 799 exceed reading speed outright.

It is a dense model, so every parameter is read for every token produced. That makes speed track memory bandwidth almost exactly — a card with twice the bandwidth generates roughly twice as fast.

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

Training it took roughly 8.6 × 10²⁰ FLOP of computation, on NVIDIA V100 — a measure of what producing the model cost, not of how fast it answers.

Around 150,000,000,000 tokens went into training it.

Step by step

How to choose a GPU for Rita-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 Rita-XLarge — around 2.0 GB at Q8_0. Capacity is the gate — a card either holds it or it does not.
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 Rita-XLarge.
03

Choose how far you will compress it

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 Rita-XLarge by squeezing it further than you would want.
04

Rank by throughput rather than spec sheet

Sort by speed to see how cards rank for Rita-XLarge. It will not match a gaming ordering — generation is bound by memory bandwidth, which is why the B200 tops it at 2,824 tok/s.
05

Read the fit column last

Tight means Rita-XLarge 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.
06

Check the card from the other side

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

Answers

Rita-XLarge — common questions

How much VRAM does Rita-XLarge need?

About 2.0 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.

Can I run Rita-XLarge on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q8_0, using about 2.0 GB and generating roughly 526 tokens per second — a comfortable fit.

Can I run Rita-XLarge on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 2.0 GB and generating roughly 322 tokens per second — a comfortable fit.

Can I run Rita-XLarge on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 2.0 GB and generating roughly 399 tokens per second — a comfortable fit.

Can I run Rita-XLarge on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 2.0 GB and generating roughly 473 tokens per second — a comfortable fit.

Is Rita-XLarge open source?

Its weights are published, so Rita-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 Rita-XLarge have?

Rita-XLarge has 1.2B parameters. "with up to 1.2 billion 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 Rita-XLarge?

Rita-XLarge was published by LightOn,Harvard University,University of Oxford, based in France, categorised as industry,Academia,Academia.

When was Rita-XLarge released?

Rita-XLarge was published in July 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 Rita-XLarge used for?

Rita-XLarge works in Biology, and is recorded as handling proteins, Protein or nucleotide language model (pLM/nLM). These are the areas it was designed around; they describe intent rather than a hard boundary.

Where can I download Rita-XLarge?

The weights for Rita-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 Rita-XLarge?

Around 8.6 × 10²⁰ FLOP, on NVIDIA V100. 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.

Can I run Rita-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 Rita-XLarge is rarely worth using. Every figure here assumes the whole model is on the card.

Would two GPUs run Rita-XLarge faster?

Two cards buy memory rather than speed. That matters for Rita-XLarge only if one card cannot hold it — 818 can, so a second adds little.

Why does the quantisation differ between cards for Rita-XLarge?

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

How accurate are these Rita-XLarge speed estimates?

They are calculated from specifications rather than measured, and each carries a range — 1,694–4,518 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.

What GPU do I need to run Rita-XLarge?

The smallest card in our catalogue that holds Rita-XLarge is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 2.0 GB, and produces roughly 30.7 tokens per second. 818 cards in total can run it.

How fast is Rita-XLarge on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 2,824 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 799 of the cards that can run Rita-XLarge clear that.

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