Granite 3.1 8B TPS calculator

Open weights IBM 8.1B parameters December 2024

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

582 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Quadro 6000

6 GB · IQ4_XS · 15.7 tok/s

Fastest card

B200

418 tok/s · 180 GB

Which GPUs can run Granite 3.1 8B?

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

582 cards match

Calculating

					Needs	Quantisation	Fit
418 tok/s 251–669 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	9.4 GB	Q8_0	Comfortable
418 tok/s 251–669 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	9.4 GB	Q8_0	Comfortable
334 tok/s 200–534 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	9.4 GB	Q8_0	Comfortable
334 tok/s 200–534 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	9.4 GB	Q8_0	Comfortable
267 tok/s 160–427 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	9.4 GB	Q8_0	Comfortable
256 tok/s 153–409 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	9.4 GB	Q8_0	Comfortable
256 tok/s 153–409 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	9.4 GB	Q8_0	Comfortable
245 tok/s 147–392 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	9.4 GB	Q8_0	Comfortable
217 tok/s 130–347 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	9.4 GB	Q8_0	Comfortable
217 tok/s 130–347 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	9.4 GB	Q8_0	Comfortable
217 tok/s 130–347 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	9.4 GB	Q8_0	Comfortable
206 tok/s 124–330 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	9.4 GB	Q8_0	Comfortable
176 tok/s 105–281 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	9.4 GB	Q8_0	Comfortable
176 tok/s 105–281 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	9.4 GB	Q8_0	Comfortable
176 tok/s 105–281 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	9.4 GB	Q8_0	Comfortable
176 tok/s 105–281 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	9.4 GB	Q8_0	Comfortable
176 tok/s 105–281 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	9.4 GB	Q8_0	Comfortable
139 tok/s 83–223 · low confidence	CMP 170HX 8 GB NVIDIA	8 GB	1,490 GB/s	Sep 2021	6.5 GB	Q5_K_M	Tight
134 tok/s 80–214 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	9.4 GB	Q8_0	Comfortable
134 tok/s 80–214 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	9.4 GB	Q8_0	Comfortable
119 tok/s 71–190 · low confidence	CMP 170HX 10 GB NVIDIA	10 GB	1,560 GB/s	Sep 2021	7.5 GB	Q6_K	Comfortable
111 tok/s 67–178 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	9.4 GB	Q8_0	Comfortable
109 tok/s 65–175 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	9.4 GB	Q8_0	Comfortable
107 tok/s 64–171 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	9.4 GB	Q8_0	Comfortable
107 tok/s 64–171 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	9.4 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: IBM
Organisation type: Industry
Country: United States of America
Published: 18 December 2024

What it does

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

Domain: Language
Task: Language modeling/generation, Question answering, Translation

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: 8.1B
Training data: 12,000,000,000,000 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: 5.8 × 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: NVIDIA H100 SXM5 80GB
Cloud vendor: IBM
Data centre: IBM's super computing cluster, Blue Vela, which is outfitted with NVIDIA H100 GPUs. This cluster provides a scalable and efficient infrastructure for training our models over thousands of GPUs.

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
Hugging Face: ibm-granite

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.

Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run Granite 3.1 8B

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 Granite 3.1 8B

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

Quadro 6000

Memory needed

5.1 GB

Fastest

418 tok/s

Granite 3.1 8B is small enough at 8.1B parameters that hardware is rarely the obstacle — 582 of the cards we track can run it, including cards several years old.

At the low end, a Quadro 6000 handles it — 6 GB, at IQ4_XS, for about 15.7 tokens per second.

The quickest result comes from a B200 at around 418 tokens per second — its 8,000 GB/s of bandwidth is what buys that.

About this model

Granite 3.1 8B was published by IBM, in United States of America, in December 2024. It comes out of industry.

It works in Language, and is recorded as doing language modeling/generation, Question answering, Translation.

The weights are published, so it can be downloaded and run on your own hardware indefinitely, offline, with no account attached. It is published under the ibm-granite organisation on Hugging Face.

How fast it runs, and why

The median result is around 23.5 tokens per second; 551 cards produce text faster than most people read it.

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.

Its internal architecture is not on file, so memory is approximated from the parameter count and marked accordingly. Expect the real figure to differ, more so at long context.

How it was trained

Producing it required around 5.8 × 10²³ FLOP of arithmetic, on NVIDIA H100 SXM5 80GB, which is a statement about the training budget rather than about inference.

The training set ran to roughly 12,000,000,000,000 tokens.

Step by step

How to choose a GPU for Granite 3.1 8B

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

Every card here has been checked against Granite 3.1 8B — around 5.1 GB at IQ4_XS. Capacity is the gate — a card either holds it or it does not.
02

Decide how long your conversations run

Set the context to what you will actually use. The cache grows with the conversation, and it is the usual reason Granite 3.1 8B stops fitting a card that seemed fine.
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 Granite 3.1 8B by squeezing it further than you would want.
04

Compare tokens per second, not specifications

Ranking by tokens per second for Granite 3.1 8B follows memory bandwidth, not core counts, which is why the B200 tops it at 418 tok/s.
05

Check the fit verdict before buying

The fit column separates cards that just manage Granite 3.1 8B from those with room to spare. Buy for the second if the context might grow.
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 Granite 3.1 8B is settled.

Answers

Granite 3.1 8B — common questions

Where can I download Granite 3.1 8B?

Its weights are published under the ibm-granite 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 Granite 3.1 8B?

Around 5.8 × 10²³ FLOP, on NVIDIA H100 SXM5 80GB. 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 Granite 3.1 8B 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 Granite 3.1 8B is rarely worth using — the nearest miss we calculate is short by 1.1 GB. Every figure here assumes the whole model is on the card.

Would two GPUs run Granite 3.1 8B faster?

A second card roughly doubles the memory available but not the generation rate. With 582 cards already able to run Granite 3.1 8B alone, the case for pairing is weak.

Why does the quantisation differ between cards for Granite 3.1 8B?

A larger card holds a more accurate copy. Across the cards that run Granite 3.1 8B, 4 compression levels are used; the floor control above pins it to one.

How accurate are these Granite 3.1 8B speed estimates?

Every figure is derived from memory bandwidth and model size, not benchmarked. That is why each is published as a range such as 251–669 tok/s on the B200 rather than a single number.

What GPU do I need to run Granite 3.1 8B?

The smallest card in our catalogue that holds Granite 3.1 8B is the Quadro 6000, with 6 GB of memory. It runs the model at IQ4_XS using about 5.1 GB, and produces roughly 15.7 tokens per second. 582 cards in total can run it.

How fast is Granite 3.1 8B on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 418 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 551 of the cards that can run Granite 3.1 8B clear that.

How much VRAM does Granite 3.1 8B need?

About 5.1 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 Granite 3.1 8B on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q5_K_M, using about 6.5 GB and generating roughly 139 tokens per second — a tight fit.

Can I run Granite 3.1 8B on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 9.4 GB and generating roughly 47.7 tokens per second — a tight fit.

Can I run Granite 3.1 8B on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 9.4 GB and generating roughly 59.1 tokens per second — a comfortable fit.

Can I run Granite 3.1 8B on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 9.4 GB and generating roughly 70.1 tokens per second — a comfortable fit.

Is Granite 3.1 8B open source?

Its weights are published, so Granite 3.1 8B 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 Granite 3.1 8B have?

Granite 3.1 8B has 8.1B parameters. 8.1B Model Architecture: Granite-3.1-8B-Base is based on a decoder-only dense transformer architecture. Core components of this architecture are: GQA and RoPE, MLP with SwiGLU, RMSNorm, and shared input/output embeddings. 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 Granite 3.1 8B?

Granite 3.1 8B was published by IBM, based in United States of America, categorised as industry.

When was Granite 3.1 8B released?

Granite 3.1 8B was published in December 2024. 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 Granite 3.1 8B used for?

Granite 3.1 8B works in Language, and is recorded as handling language modeling/generation, Question answering, Translation. 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.

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