ORNL researchers have developed Photon, a framework that accelerates the discovery of vulnerabilities in artificial intelligence models by scaling testing across the Frontier exascale supercomputer. The system uses coordinated, automated attack strategies to identify and refine weaknesses in AI models, helping improve their security and reliability in critical applications.
ORNL postdoctoral associate Jakob Bludau was named a 2026 Better Scientific Software Fellow, recognizing his leadership in advancing sustainable, high-performance scientific software and highlighting ORNL’s role in shaping the future of computational science.
ORNL is announcing the creation of the Institute for Next-Generation Data Centers, a new national institute dedicated to advancing the design, operation and integration of artificial intelligence data centers into the United States’ energy system.
Experiments conducted between 2002 and 2012 at ORNL studied 31 tin isotopes with varying numbers of neutrons to examine how neutrons affect nuclear stability and nuclear properties. The combined results contributed to identifying tin-132 as a doubly magic nucleus and improved theoretical models of nuclear structure.
ORNL has launched a novel robotic platform to rapidly analyze plant root systems as they grow, yielding AI-ready data to accelerate the development of stress-tolerant crops for new fuels, chemicals and materials. The new platform adds belowground imaging to ORNL’s Advanced Plant Phenotyping Laboratory.
Scientists at ORNL have created a new method that more than doubles computer processing speeds while using 75 percent less memory to analyze plant imaging data. The advance removes a major computational bottleneck and accelerates AI-guided discoveries for the development of high-performing crops.
In a long-running collaboration with GE Aerospace, researchers at the University of Melbourne in Australia have been steadily working to improve the performance of high-pressure turbine engines
Giri Prakash, lead for the Earth System Informatics and Data Discovery Section at ORNL, has been elevated to senior member of the Institute of Electrical and Electronics Engineers.
ORNL is beginning a four-year collaboration exploring the use of high-performance computing (HPC) to drive innovation around nonequilibrium quantum materials. This collaborative effort, Controlled Numerics for Emergent Transients in Nonequilibrium Quantum Matter will create an interdisciplinary research program to transform how scientists model and understand the complex behaviors and processes that happen when quantum materials are out of balance.
At SC25, DOE Undersecretary Dario Gil joined national labs in spotlighting the accelerating integration of AI, quantum and HPC, while ORNL delivered multiple talks and demonstrations on emerging computational technologies. ORNL teams also won Best Paper, Best Student Paper and several additional honors.
The innovative Celeritas project, led by ORNL, provides a software tool that makes sure simulations used to analyze particles can run on the fastest supercomputers, accelerating answers about the nature of the universe.
Researchers Sang-Ho Kim, An-Ping Li, Bronson Messer and Zac Ward of ORNL have been named Fellows of the American Physical Society in recognition of their outstanding impact in their respective fields.