DiffDock TPS calculator

Open weights Massachusetts Institute of Technology (MIT) 20.2M parameters October 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 · 1,821 tok/s

Fastest card

B200

167,403 tok/s · 180 GB

Which GPUs can run DiffDock?

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
167,403 tok/s 100,442–267,845 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	0.7 GB	Q8_0	Comfortable
167,403 tok/s 100,442–267,845 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	0.7 GB	Q8_0	Comfortable
133,675 tok/s 80,205–213,881 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	0.7 GB	Q8_0	Comfortable
133,675 tok/s 80,205–213,881 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	0.7 GB	Q8_0	Comfortable
106,908 tok/s 64,145–171,052 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	0.7 GB	Q8_0	Comfortable
102,325 tok/s 61,395–163,720 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	0.7 GB	Q8_0	Comfortable
102,325 tok/s 61,395–163,720 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	0.7 GB	Q8_0	Comfortable
97,931 tok/s 58,758–156,689 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	0.7 GB	Q8_0	Comfortable
86,914 tok/s 52,148–139,062 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	0.7 GB	Q8_0	Comfortable
86,914 tok/s 52,148–139,062 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	0.7 GB	Q8_0	Comfortable
86,914 tok/s 52,148–139,062 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	0.7 GB	Q8_0	Comfortable
82,446 tok/s 49,468–131,914 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	0.7 GB	Q8_0	Comfortable
70,309 tok/s 42,186–112,495 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	0.7 GB	Q8_0	Comfortable
70,309 tok/s 42,186–112,495 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	0.7 GB	Q8_0	Comfortable
70,309 tok/s 42,186–112,495 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	0.7 GB	Q8_0	Comfortable
70,309 tok/s 42,186–112,495 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	0.7 GB	Q8_0	Comfortable
70,309 tok/s 42,186–112,495 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	0.7 GB	Q8_0	Comfortable
53,535 tok/s 32,121–85,657 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	0.7 GB	Q8_0	Comfortable
53,535 tok/s 32,121–85,657 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	0.7 GB	Q8_0	Comfortable
44,613 tok/s 26,768–71,381 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	0.7 GB	Q8_0	Comfortable
43,661 tok/s 26,196–69,857 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	0.7 GB	Q8_0	Comfortable
42,688 tok/s 25,613–68,300 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	0.7 GB	Q8_0	Comfortable
42,688 tok/s 25,613–68,300 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	0.7 GB	Q8_0	Comfortable
42,688 tok/s 25,613–68,300 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	0.7 GB	Q8_0	Comfortable
42,688 tok/s 25,613–68,300 · low confidence	H100 CNX NVIDIA	80 GB	2,040 GB/s	Mar 2023	0.7 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: Massachusetts Institute of Technology (MIT)
Organisation type: Academia
Country: United States of America
Published: 4 October 2022
Authors: Gabriele Corso, Hannes Stärk, Bowen Jing, Regina Barzilay, Tommi Jaakkola

What it does

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

Domain: Biology
Task: Proteins

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: 20.2M
Training data: 4,352,000 tokens
Epochs: 850

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: 7.2 × 10¹⁹ FLOP
How it was established: Hardware

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 RTX A6000
Wall-clock time: 432 hours (18 days)

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

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: Likely
Citations: 731

Sources

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

Reference: DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking
Last updated: 25 May 2026

The extremes

The ten fastest GPUs that run DiffDock

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 DiffDock

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 C1080

Memory needed

0.7 GB

Fastest

167,403 tok/s

DiffDock is small enough at 20.2M 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 1,821 tokens per second.

The quickest result comes from a B200 at around 167,403 tokens per second — its 8,000 GB/s of bandwidth is what buys that.

Where it came from

DiffDock was published by Massachusetts Institute of Technology (MIT), in United States of America, in October 2022. academia is the category the publisher falls under.

It works in Biology, and is recorded as doing proteins.

The weights are published, so it can be downloaded and run on your own hardware indefinitely, offline, with no account attached.

Understanding the speeds

Across every card that can run it, the middle of the range is about 4,700.7 tokens per second, and 818 of them clear the ten tokens per second that roughly matches reading speed.

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.

Memory here is estimated from size rather than computed from the architecture, which is not recorded for this model — the numbers are indicative rather than exact.

What went into building it

Training it took roughly 7.2 × 10¹⁹ FLOP of computation, on NVIDIA RTX A6000 — a measure of what producing the model cost, not of how fast it answers.

Around 4,352,000 tokens went into training it.

Its inclusion criterion is sOTA improvement.

Step by step

How to choose a GPU for DiffDock

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 DiffDock — around 0.7 GB at Q8_0. Capacity is the gate — a card either holds it or it does not.
02

Match the context to your actual use

Set the context to what you will actually use. The cache grows with the conversation, and it is the usual reason DiffDock stops fitting a card that seemed fine.
03

Decide how much compression you will accept

Compression is what makes DiffDock 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.
04

Rank by throughput rather than spec sheet

Sort by speed to see how cards rank for DiffDock. It will not match a gaming ordering — generation is bound by memory bandwidth, which is why the B200 tops it at 167,403 tok/s.
05

Check the fit verdict before buying

The fit column separates cards that just manage DiffDock 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 DiffDock is settled.

Answers

DiffDock — common questions

What is DiffDock used for?

DiffDock works in Biology, and is recorded as handling proteins. These are the areas it was designed around; they describe intent rather than a hard boundary.

Where can I download DiffDock?

The weights for DiffDock 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 DiffDock?

Around 7.2 × 10¹⁹ FLOP, on NVIDIA RTX A6000. 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 DiffDock 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 DiffDock is rarely worth using. Every figure here assumes the whole model is on the card.

Would two GPUs run DiffDock faster?

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

Why does the quantisation differ between cards for DiffDock?

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

How accurate are these DiffDock 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 100,442–267,845 tok/s on the B200 rather than a single number.

What GPU do I need to run DiffDock?

The smallest card in our catalogue that holds DiffDock is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 0.7 GB, and produces roughly 1,821 tokens per second. 818 cards in total can run it.

How fast is DiffDock on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 167,403 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 DiffDock clear that.

How much VRAM does DiffDock need?

About 0.7 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 DiffDock on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q8_0, using about 0.7 GB and generating roughly 31,179 tokens per second — a comfortable fit.

Can I run DiffDock on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 0.7 GB and generating roughly 19,092 tokens per second — a comfortable fit.

Can I run DiffDock on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 0.7 GB and generating roughly 23,646 tokens per second — a comfortable fit.

Can I run DiffDock on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 0.7 GB and generating roughly 28,040 tokens per second — a comfortable fit.

Is DiffDock open source?

Its weights are published, so DiffDock 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 DiffDock have?

DiffDock has 20.2M parameters. "For determining the hyperparameters of DIFFDOCK’s score model, we trained smaller models (3.97 million parameters) that fit into 48GB of GPU RAM before scaling it up to the final model (20.24 million parameters) that was trained on four 48GB GPUs" There's a separate 4.77M "confidence model" that helps make predictions along with the score model. 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 DiffDock?

DiffDock was published by Massachusetts Institute of Technology (MIT), based in United States of America, categorised as academia.

When was DiffDock released?

DiffDock was published in October 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.

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