Aurora

Closed weights Microsoft Research 1.3B parameters May 2024

No estimate

No hardware requirements for this model

The weights for this model have not been published, so it cannot be downloaded or run on your own hardware at any size. It is reachable only through its provider, and no graphics card changes that.

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
Microsoft Research
Organisation type
Industry
Country
United States of America
Published
28 May 2024
Authors
Cristian Bodnar, Wessel P. Bruinsma, Ana Lucic, Megan Stanley, Johannes Brandstetter, Patrick Garvan, Maik Riechert, Jonathan Weyn, Haiyu Dong, Anna Vaughan, Jayesh K. Gupta, Kit Tambiratnam, Alex Archibald, Elizabeth Heider, Max Welling, Richard E. Turner, Paris Perdikaris

What it does

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

Domain
Earth science
Task
Weather forecasting
Numerical format
BF16

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
1.3B
Training data
tokens

700B tokens (see Figure 4) Total 1,219.91 TB 11,023,730 frames "All models are pretrained for 150 k steps on 32 GPUs, with a batch size of one per GPU."

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
4.5 × 10²¹ FLOP

"The model is trained using bf16 mixed precision." 312000000000000*420*32*3600*.3 = 4528742400000000000000 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 A100
Chips used
32
Wall-clock time
420 hours (17.5 days)

"32 A100 GPUs, which corresponds to approximately two and a half weeks of training" 17,5 days * 24 hours = 420

Power draw
25.3 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
Closed — provider access only
Model access
Unreleased
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

Sources

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

Reference
Aurora: A Foundation Model of the Atmosphere
Last updated
28 November 2025

What the numbers mean

Where it came from

Aurora was published by Microsoft Research, in United States of America, in May 2024. The organisation is categorised as industry.

It works in Earth science, and is recorded as doing weather forecasting.

This is a closed model: the trained values stayed with whoever produced them, and there is no local version to run.

What went into building it

Training it took roughly 4.5 × 10²¹ FLOP of computation, on NVIDIA A100 — a measure of what producing the model cost, not of how fast it answers.

Answers

Aurora — common questions

01

How much compute was used to train Aurora?

Around 4.5 × 10²¹ FLOP, on NVIDIA A100. 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.

02

What GPU do I need to run Aurora?

None. Aurora is a closed model — its weights were never published, so it cannot be downloaded or run on your own hardware at any price. It is reachable only through its provider.

03

Is Aurora open source?

No. Aurora has not had its weights published, so it exists only as a service controlled by its owner.

04

How many parameters does Aurora have?

Aurora has 1.3B 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.

05

Who created Aurora?

Aurora was published by Microsoft Research, based in United States of America, categorised as industry.

06

When was Aurora released?

Aurora was published in May 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.

07

What is Aurora used for?

Aurora works in Earth science, and is recorded as handling weather forecasting. These are the areas it was designed around; they describe intent rather than a hard boundary.

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.