Argonne Leadership Computing Facility

09/04/2026

With support from ALCF supercomputers, researchers from JPMorganChase and Argonne National Laboratory developed a new approach to evaluate the performance of the Quantum Approximate Optimization Algorithm (QAOA) at scales that were previously too computationally expensive to study.

By mapping the problem to a spin-boson system, the team replaced costly calculations with simpler simulations, providing a new way to assess the potential and limitations of quantum optimization algorithms as quantum hardware advances.

https://www.anl.gov/mcs/article/new-approach-brings-scientists-one-step-closer-to-practical-quantum-computing

09/03/2026

As part of DOE’s Genesis Mission, Argonne National Laboratory researchers and collaborators are developing the Materials Discovery Cloud, a physics-informed AI framework to predict how tiny defects affect the performance and lifetime of microelectronic devices. The effort will draw on experimental and computing capabilities at DOE user facilities, including large-scale computing resources at the ALCF.

https://www.anl.gov/article/predicting-microelectronics-performance-with-physicsinformed-artificial-intelligence

09/01/2026

How can supercomputers, automated workflows, and AI help researchers streamline fusion experiments?

The latest ALCF Service-Enabled Science session explored how D3D National Fusion Facility researchers are connecting experiments with ALCF computing resources and a digital twin to analyze plasma behavior, evaluate control strategies, and help guide future experiments. Learn more and watch the webinar: https://www.alcf.anl.gov/news/alcf-service-enabled-science-series-explores-experiment-time-computing-and-diii-d-digital-twin

Next up in the series: Learn how to build your own agentic workflow using ALCF resources at our Sept. 30 hands-on session, “Working with Jobs, Data, AI, and Agents." Reserve your spot here: https://www.alcf.anl.gov/events/service-enabled-science-jobs-data-ai-agents

08/31/2026

The Dark Energy Survey mapped 669 million galaxies to deliver one of the most detailed views yet of how the universe has evolved. As a key partner in the global collaboration, Argonne National Laboratory is helping turn the survey’s massive datasets into scientific insight through software, advanced computing, and cosmological modeling, with ALCF supercomputers supporting large-scale simulations that will help advance our understanding of dark energy and the evolution of the universe.

The Dark Energy Survey releases six years of data, delivering the most precise map yet of the universe’s large‑scale structure. Argonne is part of this international collaboration, which has shed new light on the mysteries of dark energy.

08/28/2026

With support from ALCF supercomputers, researchers from the University of Illinois Chicago developed a new physics-based method to identify a protein’s true reaction coordinates, which help drive conformational changes that can affect drug interactions. Their approach dramatically accelerates molecular simulations while enhancing their predictive power, opening new possibilities for biomedical research, including drug and enzyme design.

https://www.alcf.anl.gov/science/case-studies/understanding-rigorous-molecular-mechanism-drug-resistance-hiv-protease

08/27/2026

This summer, Reo Sze, a Ph.D. student in physics at the University of Maryland, Baltimore County (UMBC), is working at Argonne National Laboratory to enhance ChemGraph, an agentic AI framework designed to enable automated scientific workflows for chemistry and materials research. His work is focused on adding capabilities that allow ChemGraph to interpret chemical structures from images and connect them with text-based representations, as well as set up and run first-principles simulations on high-performance computing systems.

"The longer-term goal is to build an agent that can use its scientific knowledge to propose an informed initial hypothesis, test it through automated simulations, and use the resulting data to guide the next step," Sze said. "By handling repetitive setup, ex*****on, and data collection, it could allow researchers to spend more time developing ideas, interpreting results, and exploring new research directions."

Sze is part of a talented group of working on research projects at the intersection of high-performance computing, AI, and science. Follow along this summer as we highlight their projects and experiences.

Learn more about student opportunities at ALCF: https://www.alcf.anl.gov/alcf-student-opportunities

08/26/2026

STREAMLINE is a multi-institution effort using AI and high performance computing to solve the nuclear many-body problem, advancing discovery in nuclear physics, astrophysics and beyond.

08/26/2026

Join us on September 30 for a hands-on training session to learn how to use ALCF and Globus services to build agentic workflows that coordinate data, computing, and AI to carry out scientific tasks.

Part of the ALCF’s Service-Enabled Science series, the session will guide participants through managing jobs and data on ALCF systems, working with AI models through the ALCF Inference Service, and integrating models with agent environments such as Claude Code and Codex. Attendees will work through a range of examples and leave with a working agentic workflow setup they can adapt to their own research.

Register here: https://www.alcf.anl.gov/events/service-enabled-science-jobs-data-ai-agents

08/25/2026

Yu-Hsiang Lan, a Ph.D. student in computer science at the University of Illinois Urbana-Champaign, spent the summer at Argonne National Laboratory developing a scalable, high-fidelity magnetohydrodynamics (MHD) solver for fusion energy research. Working with researchers in Argonne’s Mathematics and Computer Science (MCS) Division, he is helping advance a NekRS-based solver designed to model liquid-metal flow in fusion blankets, which play a critical role in extracting energy from fusion reactors. Lan is using the ALCF’s Aurora supercomputer to test and improve the solver’s accuracy, robustness, and scalability for large-scale simulations.

“I’ve had the opportunity to run my simulations on Aurora, scaling to 10,000 nodes—nearly the entire machine—which gave me a firsthand appreciation of computing at extreme scale,” Lan said. “I’ve also had many opportunities to interact directly with researchers and domain experts and discuss my work with the scientists who will ultimately use the MHD solver I’m developing.”

Lan is part of a talented group of Argonne students contributing to projects across high-performance computing, AI, and science. Follow along with to learn more about their work and experiences.

For information on student opportunities at ALCF, visit: https://www.alcf.anl.gov/alcf-student-opportunities

08/24/2026

Join us on September 17 for “The Next Frontier: How AI Is Reshaping Discovery at Argonne,” an OutLoud public lecture exploring Argonne National Laboratory’s long history with AI and and how it is accelerating the path to scientific discovery today. ALCF Director Michael Papka will lead a discussion with Argonne researchers Ilke Arslan, Katrin Heitmann, and Ravi Madduri about the growing role of AI in science, including applications in health and cancer research, cosmology, and autonomous discovery.

Register to attend in person at Argonne or join online: https://www.anl.gov/event/the-next-frontier-how-ai-is-reshaping-discovery-at-argonne