BigGAN-deep 512x512 TPS calculator

Open weights Heriot-Watt University,DeepMind 112.7M parameters September 2018

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 · 327 tok/s

Fastest card

B200

30,066 tok/s · 180 GB

Which GPUs can run BigGAN-deep 512x512?

Set the inputs, read the answer

A longer conversation needs more memory, which can push this model off smaller cards.

Hides cards that would only fit the model by compressing it below this point.

818 cards match

Calculating
Needs Quantisation Fit
30,066 tok/s

18,039–48,105 · low confidence

B200 NVIDIA 180 GB 8,000 GB/s Jan 2024 0.8 GB Q8_0 Comfortable
30,066 tok/s

18,039–48,105 · low confidence

B300 NVIDIA 288 GB 8,000 GB/s Sep 2025 0.8 GB Q8_0 Comfortable
24,008 tok/s

14,405–38,413 · low confidence

Radeon Instinct MI350X AMD 288 GB 8,190 GB/s Jan 2025 0.8 GB Q8_0 Comfortable
24,008 tok/s

14,405–38,413 · low confidence

Radeon Instinct MI355X AMD 288 GB 8,190 GB/s Jan 2025 0.8 GB Q8_0 Comfortable
19,201 tok/s

11,520–30,721 · low confidence

Radeon Instinct MI300 AMD 128 GB 6,550 GB/s Jan 2023 0.8 GB Q8_0 Comfortable
18,378 tok/s

11,027–29,404 · low confidence

H200 NVL NVIDIA 141 GB 4,890 GB/s Nov 2024 0.8 GB Q8_0 Comfortable
18,378 tok/s

11,027–29,404 · low confidence

H200 SXM 141 GB NVIDIA 141 GB 4,890 GB/s Nov 2024 0.8 GB Q8_0 Comfortable
17,588 tok/s

10,553–28,141 · low confidence

Radeon Instinct MI325X AMD 256 GB 6,000 GB/s Oct 2024 0.8 GB Q8_0 Comfortable
15,610 tok/s

9,366–24,975 · low confidence

Radeon Instinct MI300A AMD 128 GB 5,325 GB/s Dec 2023 0.8 GB Q8_0 Comfortable
15,610 tok/s

9,366–24,975 · low confidence

Radeon Instinct MI300X AMD 192 GB 5,325 GB/s Dec 2023 0.8 GB Q8_0 Comfortable
15,610 tok/s

9,366–24,975 · low confidence

Radeon Instinct MI308X AMD 192 GB 5,325 GB/s Dec 2023 0.8 GB Q8_0 Comfortable
14,807 tok/s

8,884–23,692 · low confidence

H100 NVL 94 GB NVIDIA 94 GB 3,940 GB/s Mar 2023 0.8 GB Q8_0 Comfortable
12,628 tok/s

7,577–20,204 · low confidence

H100 PCIe 96 GB NVIDIA 96 GB 3,360 GB/s Mar 2023 0.8 GB Q8_0 Comfortable
12,628 tok/s

7,577–20,204 · low confidence

H100 SXM5 80 GB NVIDIA 80 GB 3,360 GB/s Oct 2022 0.8 GB Q8_0 Comfortable
12,628 tok/s

7,577–20,204 · low confidence

H100 SXM5 94 GB NVIDIA 94 GB 3,360 GB/s Mar 2023 0.8 GB Q8_0 Comfortable
12,628 tok/s

7,577–20,204 · low confidence

H100 SXM5 96 GB NVIDIA 96 GB 3,360 GB/s Mar 2023 0.8 GB Q8_0 Comfortable
12,628 tok/s

7,577–20,204 · low confidence

H800 SXM5 NVIDIA 80 GB 3,360 GB/s Mar 2023 0.8 GB Q8_0 Comfortable
9,615 tok/s

5,769–15,384 · low confidence

Radeon Instinct MI250 AMD 128 GB 3,280 GB/s Nov 2021 0.8 GB Q8_0 Comfortable
9,615 tok/s

5,769–15,384 · low confidence

Radeon Instinct MI250X AMD 128 GB 3,280 GB/s Nov 2021 0.8 GB Q8_0 Comfortable
8,012 tok/s

4,807–12,820 · low confidence

Data Center GPU Max 1550 Intel 128 GB 3,280 GB/s Jan 2023 0.8 GB Q8_0 Comfortable
7,841 tok/s

4,705–12,546 · low confidence

Data Center GPU Max Subsystem Intel 128 GB 3,210 GB/s Jan 2023 0.8 GB Q8_0 Comfortable
7,667 tok/s

4,600–12,267 · low confidence

A100 SXM4 80 GB NVIDIA 80 GB 2,040 GB/s Nov 2020 0.8 GB Q8_0 Comfortable
7,667 tok/s

4,600–12,267 · low confidence

A100X NVIDIA 80 GB 2,040 GB/s Jun 2021 0.8 GB Q8_0 Comfortable
7,667 tok/s

4,600–12,267 · low confidence

A800 SXM4 80 GB NVIDIA 80 GB 2,040 GB/s Aug 2022 0.8 GB Q8_0 Comfortable
7,667 tok/s

4,600–12,267 · 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
Heriot-Watt University,DeepMind
Organisation type
Academia,Industry
Country
United Kingdom of Great Britain and Northern Ireland
Published
28 September 2018
Authors
Andrew Brock, Jeff Donahue, Karen Simonyan

What it does

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

Domain
Image generation
Task
Image generation
Numerical format
FP32

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
112.7M

I used the publicly available implementation available at [1] There I loaded the biggan-deep512/1 model, and ran script [2] to compute the number of parameters [1] https://colab.research.google.com/github/tensorflow/hub/blob/master/examples/colab/biggan_generation_with_tf_hub.ipynb [2] n_params = 0 for var in module.variables: n_params += np.prod(var.shape.as_list()) pass print(n_params)

Training data
584,000,000 tokens

"To confirm that our design choices are effective for even larger and more complex and diverse datasets, we also present results of our system on a subset of JFT-300M (Sun et al., 2017). The full JFT-300M dataset contains 300M real-world images labeled with 18K categories. Since the category distribution is heavily long-tailed, we subsample the dataset to keep only images with the 8.5K most common labels. The resulting dataset contains 292M images – two orders of magnitude larger than ImageNet. …

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
1.8 × 10²¹ FLOP

3e21, estimate taken from: https://www.lesswrong.com/posts/wfpdejMWog4vEDLDg/ai-and-compute-trend-isn-t-predictive-of-what-is-happening

How it was established
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
Chips used
256
Chip-hours
12,288
Wall-clock time
48 hours

"We train on a Google TPU v3 Pod, with the number of cores proportional to the resolution: 128 for 128×128, 256 for 256×256, and 512 for 512×512. Training takes between 24 and 48 hours for most models"

Power draw
238.2 kW
Compute cost
$5,372

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

repo license is Apache: https://github.com/tensorflow/tfhub.dev/blob/master/assets/docs/deepmind/models/biggan-deep-512/1.md

How it is classified

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

Frontier model
Yes
Record confidence
Likely
Citations
6,101

Sources

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

Reference
Large Scale GAN Training for High Fidelity Natural Image Synthesis
Last updated
25 May 2026

The extremes

The ten fastest GPUs that run BigGAN-deep 512x512

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.

  1. 01 B300 288 GB · 8,000 GB/s · Q8_0 30,066 tok/s
  2. 02 B200 180 GB · 8,000 GB/s · Q8_0 30,066 tok/s
  3. 03 Radeon Instinct MI350X 288 GB · 8,190 GB/s · Q8_0 24,008 tok/s
  4. 04 Radeon Instinct MI355X 288 GB · 8,190 GB/s · Q8_0 24,008 tok/s
  5. 05 Radeon Instinct MI300 128 GB · 6,550 GB/s · Q8_0 19,201 tok/s
  6. 06 H200 NVL 141 GB · 4,890 GB/s · Q8_0 18,378 tok/s
  7. 07 H200 SXM 141 GB 141 GB · 4,890 GB/s · Q8_0 18,378 tok/s
  8. 08 Radeon Instinct MI325X 256 GB · 6,000 GB/s · Q8_0 17,588 tok/s
  9. 09 Radeon Instinct MI300A 128 GB · 5,325 GB/s · Q8_0 15,610 tok/s
  10. 10 Radeon Instinct MI300X 192 GB · 5,325 GB/s · Q8_0 15,610 tok/s

The smallest GPUs that still run BigGAN-deep 512x512

The cheapest route in, by memory capacity. A tight fit runs the model but leaves nothing spare for a longer conversation.

  1. 01 GeForce RTX 4010 4 GB · needs 0.8 GB · Q8_0 · comfortable 361 tok/s
  2. 02 RTX A400 4 GB · needs 0.8 GB · Q8_0 · comfortable 361 tok/s
  3. 03 RTX 500 Mobile Ada Generation 4 GB · needs 0.8 GB · Q8_0 · comfortable 481 tok/s
  4. 04 GeForce RTX 3050 A Mobile 4 GB · needs 0.8 GB · Q8_0 · comfortable 722 tok/s
  5. 05 Jetson Orin Nano 4 GB 4 GB · needs 0.8 GB · Q8_0 · comfortable 128 tok/s
  6. 06 Radeon RX 6450M 4 GB · needs 0.8 GB · Q8_0 · comfortable 375 tok/s
  7. 07 Radeon RX 6550M 4 GB · needs 0.8 GB · Q8_0 · comfortable 422 tok/s
  8. 08 Radeon RX 6550S 4 GB · needs 0.8 GB · Q8_0 · comfortable 375 tok/s
  9. 09 Arc A310 4 GB · needs 0.8 GB · Q8_0 · comfortable 303 tok/s
  10. 10 Arc Pro A30M 4 GB · needs 0.8 GB · Q8_0 · comfortable 313 tok/s

What the numbers mean

The hardware side

Minimum card

Tesla C1080

Memory needed

0.8 GB

Fastest

30,066 tok/s

BigGAN-deep 512x512 is small enough at 112.7M parameters that hardware is rarely the obstacle — 818 of the cards we track can run it, including cards several years old.

At the low end, a Tesla C1080 handles it — 4 GB, at Q8_0, for about 327 tokens per second.

The quickest result comes from a B200 at around 30,066 tokens per second — its 8,000 GB/s of bandwidth is what buys that.

About this model

BigGAN-deep 512x512 was published by Heriot-Watt University,DeepMind, in United Kingdom of Great Britain and Northern Ireland, in September 2018. The organisation is categorised as academia,Industry.

It works in Image generation, and is recorded as doing image generation.

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 844.2 tokens per second, and 818 exceed reading speed outright.

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.

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.

What went into building it

Producing it required around 1.8 × 10²¹ FLOP of arithmetic, on Google TPU v3, which is a statement about the training budget rather than about inference.

Around 584,000,000 tokens went into training it.

Step by step

How to choose a GPU for BigGAN-deep 512x512

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

  1. 01

    Check what it needs before anything else

    Look at what BigGAN-deep 512x512 actually needs — around 0.8 GB at Q8_0. No amount of processing power compensates for a card that cannot hold it.

  2. 02

    Match the context to your actual use

    Longer conversations cost memory on top of what the weights need. Move the slider to your real working length before trusting any row for BigGAN-deep 512x512.

  3. 03

    Set a quality floor

    Compression is what makes BigGAN-deep 512x512 fit smaller cards, at some cost in accuracy — Q8_0 on the smallest card that fits. A minimum quality removes the ones that go too far.

  4. 04

    Rank by throughput rather than spec sheet

    Sort by speed to see how cards rank for BigGAN-deep 512x512. It will not match a gaming ordering — generation is bound by memory bandwidth, which is why the B200 tops it at 30,066 tok/s.

  5. 05

    Check the fit verdict before buying

    The fit column separates cards that just manage BigGAN-deep 512x512 from those with room to spare. Buy for the second if the context might grow.

  6. 06

    See what else that card runs

    Each card page repeats this sweep for every model we hold. It answers what else the hardware is good for, beyond BigGAN-deep 512x512.

Answers

BigGAN-deep 512x512 — common questions

01

Is BigGAN-deep 512x512 open source?

Its weights are published, so BigGAN-deep 512x512 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.

02

How many parameters does BigGAN-deep 512x512 have?

BigGAN-deep 512x512 has 112.7M parameters. I used the publicly available implementation available at [1] There I loaded the biggan-deep512/1 model, and ran script [2] to compute the number of parameters [1] https://colab.research.google.com/github/tensorflow/hub/blob/master/examples/colab/biggan_generation_with_tf_hub.ipynb [2] n_params = 0 for var in module.variables: n_params += np.prod(var.shape.as_list()) pass print(n_params). 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.

03

Who created BigGAN-deep 512x512?

BigGAN-deep 512x512 was published by Heriot-Watt University,DeepMind, based in United Kingdom of Great Britain and Northern Ireland, categorised as academia,Industry.

04

When was BigGAN-deep 512x512 released?

BigGAN-deep 512x512 was published in September 2018. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

05

What is BigGAN-deep 512x512 used for?

BigGAN-deep 512x512 works in Image generation, and is recorded as handling image generation. These are the areas it was designed around; they describe intent rather than a hard boundary.

06

Where can I download BigGAN-deep 512x512?

The weights for BigGAN-deep 512x512 are published, though we do not hold a repository link for it. This site calculates hardware requirements rather than hosting model files.

07

How much compute was used to train BigGAN-deep 512x512?

Around 1.8 × 10²¹ 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.

08

Can I run BigGAN-deep 512x512 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 BigGAN-deep 512x512 is rarely worth using. Every figure here assumes the whole model is on the card.

09

Would two GPUs run BigGAN-deep 512x512 faster?

A second card roughly doubles the memory available but not the generation rate. With 818 cards already able to run BigGAN-deep 512x512 alone, the case for pairing is weak.

10

Why does the quantisation differ between cards for BigGAN-deep 512x512?

Each card is shown running the least-compressed copy it can hold, and BigGAN-deep 512x512 appears at 1 different compression levels across the cards that fit it. Bigger cards get the more accurate version.

11

How accurate are these BigGAN-deep 512x512 speed estimates?

They are calculated from specifications rather than measured, and each carries a range — 18,039–48,105 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.

12

What GPU do I need to run BigGAN-deep 512x512?

The smallest card in our catalogue that holds BigGAN-deep 512x512 is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 0.8 GB, and produces roughly 327 tokens per second. 818 cards in total can run it.

13

How fast is BigGAN-deep 512x512 on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 30,066 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 818 of the cards that can run BigGAN-deep 512x512 clear that.

14

How much VRAM does BigGAN-deep 512x512 need?

About 0.8 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.

15

Can I run BigGAN-deep 512x512 on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q8_0, using about 0.8 GB and generating roughly 5,600 tokens per second — a comfortable fit.

16

Can I run BigGAN-deep 512x512 on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 0.8 GB and generating roughly 3,429 tokens per second — a comfortable fit.

17

Can I run BigGAN-deep 512x512 on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 0.8 GB and generating roughly 4,247 tokens per second — a comfortable fit.

18

Can I run BigGAN-deep 512x512 on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 0.8 GB and generating roughly 5,036 tokens per second — a comfortable fit.

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