MACE-MP-0

Open weights University of Cambridge,Federal Institute of Materials Research and Testing (BAM),NERSC, Lawrence Berkeley National Laboratory,University of British Columbia (UBC),Friedrich Schiller University Jena,University of Bayreuth,Fritz Haber Institute of the Max Planck Society,U. S. Naval Research Laboratory,Chemix,Daresbury Laboratory,BASF,University of South Carolina,University of Stuttgart,Uppsala University,Newcastle University,Technical University of Denmark,Aix-Marseille Université,University of Warwick,University of California Los Angeles (UCLA),InstaDeep,University of California (UC) Berkeley March 2024

No estimate

No hardware requirements for this model

This model's weights are open, but no parameter count has been published for it. Every memory and speed figure starts from that number, so we would rather show nothing than a fabricated estimate.

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
University of Cambridge,Federal Institute of Materials Research and Testing (BAM),NERSC, Lawrence Berkeley National Laboratory,University of British Columbia (UBC),Friedrich Schiller University Jena,University of Bayreuth,Fritz Haber Institute of the Max Planck Society,U. S. Naval Research Laboratory,Chemix,Daresbury Laboratory,BASF,University of South Carolina,University of Stuttgart,Uppsala University,Newcastle University,Technical University of Denmark,Aix-Marseille Université,University of Warwick,University of California Los Angeles (UCLA),InstaDeep,University of California (UC) Berkeley
Organisation type
Academia,Academia,Government,Academia,Academia,Academia,Academia,Government,Industry,Academia,Industry,Academia,Academia,Academia,Academia,Academia,Academia,Academia,Academia,Industry,Academia
Country
United Kingdom of Great Britain and Northern Ireland, Germany, United States of America, Canada, Sweden, Denmark, France
Published
1 March 2024
Authors
Ilyes Batatia, Philipp Benner, Yuan Chiang, Alin M. Elena, Dávid P. Kovács, Janosh Riebesell, Xavier R. Advincula, Mark Asta, Matthew Avaylon, William J. Baldwin, Fabian Berger, Noam Bernstein, Arghya Bhowmik, Samuel M. Blau, Vlad Cărare, James P. Darby, Sandip De, Flaviano Della Pia, Volker L. Deringer, Rokas Elijošius, Zakariya El-Machachi, Fabio Falcioni, Edvin Fako, Andrea C. Ferrari, Annalena…

What it does

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

Domain
Materials science
Task
Molecular simulation

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.

Training data
tokens
Epochs
200

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
8.8 × 10²⁰ FLOP

312000000000000*2600*3600*0.3 = 8.76096e+20

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
80
Chip-hours
2,600
Power draw
63.3 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
Open — downloadable
Model access
Open weights (unrestricted)
Training code
Open source

MIT License https://github.com/ACEsuit/mace-mp https://github.com/ACEsuit/mace/

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
A foundation model for atomistic materials chemistry
Last updated
28 November 2025

What the numbers mean

Where it came from

MACE-MP-0 was published by University of Cambridge,Federal Institute of Materials Research and Testing (BAM),NERSC, Lawrence Berkeley National Laboratory,University of British Columbia (UBC),Friedrich Schiller University Jena,University of Bayreuth,Fritz Haber Institute of the Max Planck Society,U. S. Naval Research Laboratory,Chemix,Daresbury Laboratory,BASF,University of South Carolina,University of Stuttgart,Uppsala University,Newcastle University,Technical University of Denmark,Aix-Marseille Université,University of Warwick,University of California Los Angeles (UCLA),InstaDeep,University of California (UC) Berkeley, in United Kingdom of Great Britain and Northern Ireland, in March 2024. It comes out of academia,Academia,Government,Academia,Academia,Academia,Academia,Government,Industry,Academia,Industry,Academia,Academia,Academia,Academia,Academia,Academia,Academia,Academia,Industry,Academia.

It works in Materials science, and is recorded as doing molecular simulation.

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

What went into building it

Producing it required around 8.8 × 10²⁰ FLOP of arithmetic, on NVIDIA A100, which is a statement about the training budget rather than about inference.

Answers

MACE-MP-0 — common questions

01

When was MACE-MP-0 released?

MACE-MP-0 was published in March 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.

02

What is MACE-MP-0 used for?

MACE-MP-0 works in Materials science, and is recorded as handling molecular simulation. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

03

Where can I download MACE-MP-0?

The weights for MACE-MP-0 are published, though we do not hold a repository link for it. This site calculates hardware requirements rather than hosting model files.

04

How much compute was used to train MACE-MP-0?

Around 8.8 × 10²⁰ 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.

05

What GPU do I need to run MACE-MP-0?

We cannot say. MACE-MP-0 has open weights, but no parameter count has been published for it, and every memory and speed calculation starts from that number. We would rather show nothing than a fabricated estimate.

06

Is MACE-MP-0 open source?

Its weights are published, so MACE-MP-0 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.

07

How many parameters does MACE-MP-0 have?

No parameter count has been published for MACE-MP-0, which is why no memory or speed figure appears on this page.

08

Who created MACE-MP-0?

MACE-MP-0 was published by University of Cambridge,Federal Institute of Materials Research and Testing (BAM),NERSC, Lawrence Berkeley National Laboratory,University of British Columbia (UBC),Friedrich Schiller University Jena,University of Bayreuth,Fritz Haber Institute of the Max Planck Society,U. S. Naval Research Laboratory,Chemix,Daresbury Laboratory,BASF,University of South Carolina,University of Stuttgart,Uppsala University,Newcastle University,Technical University of Denmark,Aix-Marseille Université,University of Warwick,University of California Los Angeles (UCLA),InstaDeep,University of California (UC) Berkeley, based in United Kingdom of Great Britain and Northern Ireland, categorised as academia,Academia,Government,Academia,Academia,Academia,Academia,Government,Industry,Academia,Industry,Academia,Academia,Academia,Academia,Academia,Academia,Academia,Academia,Industry,Academia.

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.