Andrew Fowler
Andrew Fowler applies machine learning to materials design and atomistic simulation. His early work studies uncertainty in learned electron densities to accelerate density-functional-theory calculations. At Microsoft Research AI for Science, he coauthors MatterGen, MatterSim, and MatterSim-MT for property-guided design and simulation. As of October 2026, he is a Senior Researcher with interests in materials structure prediction and discovery.
- Now at
- Microsoft Research
- Works
- 9
- Citations
- 718
- h-index
- 5
Publication numbers from OpenAlex, as of 18 October 2026.
Timeline
-
2018–2019
Uncertainty in density functional theory
In a 2018–2019 study, he uses uncertainty estimates to distinguish accurate from inaccurate predicted electron densities and initializes density functional theory calculations with trusted predictions to speed convergence.
-
2023
MatterGen for property-guided design
In 2023, he and colleagues at Microsoft Research AI for Science introduce MatterGen, a diffusion model that generates stable materials under chemistry, symmetry, and property constraints.
-
May 2024
A deep-learning atomistic model
In May 2024, he coauthors MatterSim, a deep-learning model trained on first-principles computations to simulate materials and predict properties across elements, temperatures, and pressures.
-
2025
Equal credit for MatterGen
Nature publishes the MatterGen study in 2025; Fowler is among seven equal contributors who conceive the study, implement its methods, perform computational experiments, and write the paper.
-
May 2026
The MatterSim-MT release
In May 2026, he coauthors Microsoft's MatterSim-MT announcement, which introduces a multi-task foundation model for materials simulation and property characterization.
-
As of October 2026
Senior Researcher at Microsoft
As of October 2026, he is a Senior Researcher at Microsoft Research AI for Science, with interests in machine-learning methods for structure prediction and materials discovery.
Most-cited works
- A generative model for inorganic materials design
- MatterSim: A Deep Learning Atomistic Model Across Elements, Temperatures and Pressures
- MatterGen: a generative model for inorganic materials design
- Learning models for electron densities with Bayesian regression
- Managing uncertainty in data-derived densities to accelerate density functional theory
Colleagues
Sources and credits
Photo: We have no published photo of Andrew Fowler yet. Send us a free-licensed photo.
- Managing uncertainty in data-derived densities to accelerate density functional theory
- MatterGen: Property-guided materials design
- MatterGen: a generative model for inorganic materials design
- MatterSim: A Deep Learning Atomistic Model Across Elements, Temperatures and Pressures
- A generative model for inorganic materials design
- MatterSim-MT: A multi-task foundation model for in silico materials characterization
- Advancing AI for materials with MatterSim: experimental synthesis, faster simulation, and multi-task models
- Andrew Fowler