Utilizing the Full Power of Empire AI – 2027 supports researchers developing and using foundational machine learning models on multinode GPU clusters to solve fundamental mathematics and basic science problems.
Funder: Simons Foundation
Due Dates: December 1, 2026, at 12 p.m. (noon) Eastern time (Full application)
Funding Amounts: $6 million USD total across six to 12 awards over two years; no individual award maximum
Summary: Supports foundational machine learning research using Empire AI’s multinode GPU clusters to solve fundamental mathematics and basic science problems.
Key Information: The lead PI must hold a tenure-track or tenured faculty position at an eligible Empire AI consortium institution.
The Simons Foundation’s Utilizing the Full Power of Empire AI program supports ambitious two-year research programs drawing on Empire AI’s advanced computational power. Eligible fields are computational astrophysics and physics, computational biology, mathematics, neuroscience and computational physical chemistry, including work at their intersections. Projects must develop and use foundational machine learning models to address problems of fundamental importance.
Representative research areas include computational biophysics, ecology and evolution, plant biology, optimal transport, multiscale physics, fluid dynamics and numerical general relativity. All proposals must explain how the research will use Empire AI’s multinode GPU clusters. Data curation and dataset creation may be supported exceptionally, provided the proposers use the resulting data to solve scientific problems within the same project.