The ALCF user community continually pushes the boundaries of scientific computing, producing groundbreaking studies in areas ranging from physics and engineering to biology and materials science. Below, we highlight some of the recent results published by ALCF users in peer-reviewed journals.

Ab Initio Investigation of a Hypersonic Double Cone Experiment

ALCF principal investigator (PI): Maninder Grover, Air Force Research Laboratory/University of Dayton Research Institute

As detailed in a paper published in Science Advances, researchers used ALCF resources to perform direct molecular simulations (DMS) of a Mach 8.2 oxygen flow over a double cone geometry. This configuration generates a flow with thermal and chemical nonequilibrium, which are common attributes of hypersonic flight. The team’s DMS approach leverages quantum mechanically derived interaction potentials to model molecular collisions, providing a benchmark for lower-fidelity models. The study’s large-scale simulations offer new insights into shock-shock interactions and the molecular-level mechanisms that drive hypersonic flows.

Enhanced Sampling of Protein Conformational Changes via True Reaction Coordinates from Energy Relaxation

ALCF PI: Ao Ma, University of Illinois Chicago

In a study published in Nature Communications, researchers leveraged ALCF supercomputers to develop a novel approach for accelerating protein conformational changes in molecular dynamics simulations. By identifying true reaction coordinates based on energy relaxation, they achieved a 10⁵- to 10¹⁵-fold speedup in simulations of protein conformational changes and ligand dissociation. This method allows for the generation of unbiased, natural trajectories that effectively predict transition pathways, overcoming limitations in traditional enhanced sampling and unlocking access to a broader range of protein functional processes in molecular dynamics simulations.

High-Energy LiNiO₂ Li Metal Batteries Enabled by Hybrid Electrolyte Consisting of Ionic Liquid and Weakly Solvating Fluorinated Ether

ALCF PI: Wei Jiang, Argonne National Laboratory

In a paper published in Advanced Science, researchers used ALCF computing resources to aid in the development of a hybrid electrolyte combining a highly fluorinated ionic liquid with a weakly solvating fluorinated ether. This electrolyte stabilized high-energy lithium-metal (Li0) batteries with LiNiO₂ (Ni = 100%) cathodes, enabling the batteries to achieve near-theoretical capacity (up to 249 mAh g⁻¹) for over 300 cycles while maintaining 78.6% capacity retention and preventing unwanted morphological changes in both electrodes. Molecular dynamics simulations and density functional theory calculations revealed that the fluorinated ether significantly alters the Li⁺ solvation environment, which in turn stabilizes the electrode/electrolyte interfaces. These findings open new avenues for designing next-generation electrolytes and interphases for high-energy-density batteries.

Impact of Varying BLAS Precision on DCMESH

ALCF PI: Aiichiro Nakano, University of Southern California

Published in the SC24 Conference Proceedings, this paper explores the use of various BLAS precision modes (BF16, TF32, and Complex 3M) to improve the performance of DCMESH, a framework for studying light-matter interactions, on the Intel GPUs that power Aurora. By switching between these precision modes, the team achieved up to a 3.91x speedup for individual BLAS calls while retaining accuracy in key metrics such as electron excitation, current density, and kinetic energy. The approach, which requires no source code changes (only adjustments to environment variables), demonstrates the potential for optimizations in large-scale quantum molecular simulations and can be used to enhance other HPC workloads that spend a significant amount of time in BLAS calls.

If your team has a recent publication that used ALCF resources, please let us know by contacting us at pubs@alcf.anl.gov.