Poro 34B TPS calculator

Open weights High-Performance Language Technologies (HPLT),University of Turku 34.2B parameters December 2023

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

132 of 818 cards that can run it

Smallest card that fits

RTX A4500

20 GB · Q3_K_M · 21.4 tok/s

Fastest card

B200

99.1 tok/s · 180 GB

Which GPUs can run Poro 34B?

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

132 cards match

Calculating

					Needs	Quantisation	Fit
99.1 tok/s 59–159 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	37.3 GB	Q8_0	Comfortable
99.1 tok/s 59–159 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	37.3 GB	Q8_0	Comfortable
79.1 tok/s 47–127 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	37.3 GB	Q8_0	Comfortable
79.1 tok/s 47–127 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	37.3 GB	Q8_0	Comfortable
63.3 tok/s 38–101 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	37.3 GB	Q8_0	Comfortable
60.6 tok/s 36–97 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	37.3 GB	Q8_0	Comfortable
60.6 tok/s 36–97 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	37.3 GB	Q8_0	Comfortable
58.0 tok/s 35–93 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	37.3 GB	Q8_0	Comfortable
51.4 tok/s 31–82 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	37.3 GB	Q8_0	Comfortable
51.4 tok/s 31–82 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	37.3 GB	Q8_0	Comfortable
51.4 tok/s 31–82 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	37.3 GB	Q8_0	Comfortable
48.8 tok/s 29–78 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	37.3 GB	Q8_0	Comfortable
41.6 tok/s 25–67 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	37.3 GB	Q8_0	Comfortable
41.6 tok/s 25–67 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	37.3 GB	Q8_0	Comfortable
41.6 tok/s 25–67 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	37.3 GB	Q8_0	Comfortable
41.6 tok/s 25–67 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	37.3 GB	Q8_0	Comfortable
41.6 tok/s 25–67 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	37.3 GB	Q8_0	Comfortable
41.4 tok/s 25–66 · low confidence	DRIVE A100 PROD NVIDIA	32 GB	1,870 GB/s	May 2020	25.4 GB	Q5_K_M	Tight
41.4 tok/s 25–66 · low confidence	GRID A100A NVIDIA	32 GB	1,870 GB/s	May 2020	25.4 GB	Q5_K_M	Tight
39.6 tok/s 24–63 · low confidence	GeForce RTX 5090 NVIDIA	32 GB	1,790 GB/s	Jan 2025	25.4 GB	Q5_K_M	Tight
39.6 tok/s 24–63 · low confidence	GeForce RTX 5090 D NVIDIA	32 GB	1,790 GB/s	Jan 2025	25.4 GB	Q5_K_M	Tight
38.3 tok/s 23–61 · low confidence	GeForce RTX 5090 D V2 NVIDIA	24 GB	1,340 GB/s	Aug 2025	21.4 GB	Q4_K_M	Tight
34.9 tok/s 21–56 · low confidence	A30X NVIDIA	24 GB	1,220 GB/s	Apr 2021	21.4 GB	Q4_K_M	Tight
31.7 tok/s 19–51 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	37.3 GB	Q8_0	Comfortable
31.7 tok/s 19–51 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	37.3 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: High-Performance Language Technologies (HPLT),University of Turku
Organisation type: Research collective,Academia
Country: Multinational, Finland
Published: 14 December 2023
Authors: Risto Luukkonen, Jonathan Burdge, Elaine Zosa, Aarne Talman, Ville Komulainen, Väinö Hatanpää, Peter Sarlin, Sampo Pyysalo

What it does

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

Domain: Language
Task: Code generation, Language modeling/generation

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: 34.2B
Training data: 1,000,000,000,000 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: 2.1 × 10²³ 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: AMD Radeon Instinct MI250X
Chips used: 512
Power draw: 507.5 kW

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
Hugging Face: LumiOpen

How it is classified

Labels the source dataset applies when tracking notable models, and how confident it is in the entry.

Likely above 10²³ FLOP: Yes
Record confidence: Confident

Sources

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

Reference: Poro 34B and the Blessing of Multilinguality
Last updated: 28 November 2025

The extremes

The ten fastest GPUs for Poro 34B

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 Poro 34B

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

RTX A4500

Memory needed

17.4 GB

Fastest

99.1 tok/s

With 34.2B parameters, Poro 34B lands in the range a serious desktop card can handle once the weights are compressed. 132 of the cards we track can run it.

The entry point is the RTX A4500: 20 GB of memory, Q3_K_M compression, roughly 21.4 tokens per second.

At the other end, a B200 generates roughly 99.1 tokens per second on it, on the strength of 8,000 GB/s of memory bandwidth.

About this model

Poro 34B was published by High-Performance Language Technologies (HPLT),University of Turku, in Multinational, in December 2023. research collective,Academia is the category the publisher falls under.

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

Published weights mean the model runs on your machine rather than someone else's, which is what makes the hardware question below answerable at all. It is published under the LumiOpen organisation on Hugging Face.

How fast it runs, and why

Half the cards that hold it manage more than 21.0 tokens per second, and 103 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.

Training and provenance

Producing it required around 2.1 × 10²³ FLOP of arithmetic, on AMD Radeon Instinct MI250X, which is a statement about the training budget rather than about inference.

The training set ran to roughly 1,000,000,000,000 tokens.

Step by step

How to choose a GPU for Poro 34B

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

The table lists every card that can hold Poro 34B — around 17.4 GB at Q3_K_M. That figure, not the card's headline performance, is what decides whether it runs.
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 Poro 34B.
03

Decide how much compression you will accept

Compression is what makes Poro 34B fit smaller cards, at some cost in accuracy — Q3_K_M on the smallest card that fits. A minimum quality removes the ones that go too far.
04

Sort by speed

Ranking by tokens per second for Poro 34B follows memory bandwidth, not core counts, which is why the B200 tops it at 99.1 tok/s.
05

Check the fit verdict before buying

The fit column separates cards that just manage Poro 34B from those with room to spare. Buy for the second if the context might grow.
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. Worth a look before buying for Poro 34B alone — a card is usually bought for more than one model.

Answers

Poro 34B — common questions

What GPU do I need to run Poro 34B?

The smallest card in our catalogue that holds Poro 34B is the RTX A4500, with 20 GB of memory. It runs the model at Q3_K_M using about 17.4 GB, and produces roughly 21.4 tokens per second. 132 cards in total can run it.

How fast is Poro 34B on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 99.1 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 103 of the cards that can run Poro 34B clear that.

How much VRAM does Poro 34B need?

About 17.4 GB at Q3_K_M 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 Poro 34B on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q4_K_M, using about 21.4 GB and generating roughly 38.3 tokens per second — a tight fit.

Is Poro 34B open source?

Its weights are published, so Poro 34B 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 Poro 34B have?

Poro 34B has 34.2B parameters. https://huggingface.co/LumiOpen/Poro-34B. 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 Poro 34B?

Poro 34B was published by High-Performance Language Technologies (HPLT),University of Turku, based in Multinational, categorised as research collective,Academia.

When was Poro 34B released?

Poro 34B was published in December 2023. 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 Poro 34B used for?

Poro 34B works in Language, and is recorded as handling code generation, Language modeling/generation. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Where can I download Poro 34B?

Its weights are published under the LumiOpen organisation on Hugging Face. We do not host model files — this site calculates what hardware is needed to run them.

How much compute was used to train Poro 34B?

Around 2.1 × 10²³ FLOP, on AMD Radeon Instinct MI250X. 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 Poro 34B 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 — the nearest miss we calculate is short by 7.0 GB. Our figures for Poro 34B assume it is fully resident.

Would two GPUs run Poro 34B faster?

A second card roughly doubles the memory available but not the generation rate. With 132 cards already able to run Poro 34B alone, the case for pairing is weak.

Why does the quantisation differ between cards for Poro 34B?

A larger card holds a more accurate copy. Across the cards that run Poro 34B, 5 compression levels are used; the floor control above pins it to one.

How accurate are these Poro 34B speed estimates?

They are calculated from specifications rather than measured, and each carries a range — 59–159 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.

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.

All AI models