Projects

This project is using exascale simulations and first-principles calculations to generate an open-source, high-accuracy database of gas-phase reaction rates and thermochemical properties, enabling more predictive modeling of chemical kinetics across diverse applications.

This project is carrying out high-fidelity computational simulations to evaluate modeling approaches and provide fundamental insights into fluid-dynamics phenomena in next-generation narrow-body aircraft engines, providing insights to guide accurate and cost-effective engine design.

A project team lead by researchers from Carnegie Mellon University will combine machine learning with iterative excited-state simulations to efficiently discover and predict high-performance crystalline organic semiconductors, accelerating the design of next-generation optoelectronic devices with minimal computational cost.

This project will use lattice QCD and leadership-class computing to compute quark and gluon generalized parton distributions of nucleons and pions, providing ab initio insights into proton and neutron structure and supporting experiments at Jefferson National Laboratory and the future Electron-Ion Collider.

This project will develop multimodal AI foundation models to rapidly predict and design new materials, accelerating discovery across applications like energy storage and electronics by efficiently screening millions of candidate materials.

This project will apply benchmark-accuracy quantum Monte Carlo methods to predict and understand the properties of quantum materials, providing reference data for machine-learning approaches and advancing theoretical tools for designing next-generation quantum devices.

This project will use first-principles simulations with physical quark masses to map the QCD phase diagram, compute the equation of state at nonzero density, and identify signatures of the critical endpoint and chiral critical behavior. These results will guide experiments and help advance understanding of high-density strongly interacting matter.

This project will use advanced ab initio quantum many-body methods and high-performance computing to predict nuclear structure, reactions, and fundamental interactions, enabling simulations beyond current experimental reach. It will advance the science of nuclei at major laboratories, driving discoveries in nuclear physics, astrophysics, and fundamental symmetries.

In this project, the team will develop community software that integrates high-accuracy quantum Monte Carlo calculations with multiscale and machine-learning models to surpass density functional theory in predictive power. This will target correlated materials problems including high-pressure hydrogen, advanced battery materials, and charge density waves, while generating benchmark-quality data for the broader materials community.

This project is developing numerical techniques on ALCF supercomputers to predict the performance and optimal parameters of the Quantum Approximate Optimization Algorithm (QAOA) for large combinatorial problems, providing insights into its potential for achieving quantum advantage in optimization.

This project performs high-precision numerical calculations within the Standard Model using lattice quantum chromodynamics to determine properties of hadrons and key particle decays. It seeks evidence of previously unknown physical processes by comparing theory with accelerator experiments, including measurements of the muon’s anomalous magnetic moment.

This project combines advanced sampling methods, electronic-structure calculations, and machine learning to model the dynamic ensembles of catalytic interfaces under realistic reaction conditions, enabling the predictive design of catalysts for applications such as chemical manufacturing and fuel production.

This project performs the first lattice QCD calculation of transverse-momentum–dependent parton distributions for a transversely polarized nucleon, enabling 3D imaging of its transverse spin structure. The results will provide essential theoretical input for hadron physics experiments.

This project is using exascale computing, quantum-accurate simulations, and a specialized AI foundation model to predict how RNA switches and metal-based drugs interact at the quantum level, accelerating the discovery of safer antibiotics, more effective cancer therapies, and personalized medicines.

This project advances first-principles many-body Green’s function methods to understand and predict excited-state phenomena in quantum materials. By developing and applying GW-based computational approaches on exascale supercomputers, it investigates quasiparticles, excitons, electron–phonon interactions, and nonequilibrium processes that govern optical properties and emergent quantum phases.

This project investigates how radiation damage affects candidate plasma-facing materials for future fusion reactors. Using atomistic molecular dynamics simulations, it examines ultrafast, non-equilibrium processes that govern material degradation, aiming to improve lifetime predictions and address a key barrier to commercially viable inertial and magnetic confinement fusion energy systems.

This project combines first-principles and AI-accelerated multiscale simulations with DOE experimental data to uncover how light drives electronic and atomic dynamics in quantum materials, enabling ultrafast ferroelectric switching and energy-efficient synthesis of high-temperature ceramics.

This project is using large-scale molecular dynamics simulations and AI-based structure predictions to model how inner-ear tip links transmit mechanical forces to ion channels, revealing the molecular mechanics of hearing and how mutations can lead to inherited deafness.

This project leverages advanced 3D radiation-hydrodynamic simulations and leadership-class supercomputers to produce the most comprehensive long-duration models of core-collapse supernova explosions to date, illuminating their mechanisms, observable signatures, and lasting impact on neutron stars, black holes, and the chemical evolution of the Universe.

This project is inventing ways to train and run giant science AI models far faster and with much less energy—using smart compression on supercomputers—so they’re cheaper to build, easier to use, and don’t crowd out other critical research.