Resolving how nanoparticles come together is important for industry and environmental remediation. New work predicts nanoparticle aggregation behavior across a wide range of scales for the first time.

Researchers created a methodology to design custom network topologies that improve the execution time for scientific applications.

Researchers from PNNL, CSET, and Google DeepMind delved into the potential risks of artificial intelligence in a roundtable discussion.

A poem inspired by radioactive tank waste—“Can a Scientist Dream it Alone?”—was awarded first place in the Department of Energy’s Poetry of Science Art Contest.

Pasco School District teachers join PNNL for an immersive teaching experience.

The Human Factors Symposium took place at Discovery Hall at PNNL in May 2023. Fifty-seven attendees participated in the three-day event representing 15 different institutions.

PNNL researchers present challenges and opportunities for fusing graph neural networks with deep reinforcement learning.

IDREAM research shows that keeping only the most important two- and three-body terms in reactive force fields can decrease computational cost by one order of magnitude, while preserving satisfactory accuracy.

Computational chemistry libraries from NWChem, NWChemEX, SPEC, and more are now available on Azure Quantum Elements.

PNNL researchers outline open problems in different scientific areas in the Journal of Combinatorics.

PNNL and The University of Texas at El Paso expand their premier partnership to include the Cyber Halo Innovation Research Program.

Under Secretary for Science and Innovation at the Department of Energy Geri Richmond visited PNNL.

Irradiated boehmite nanoplatelets form smaller aggregates and have a rougher surface than non-irradiated platelets.

Advances in understanding the nature of the chemical bond are featured in this special issue of The Journal of Chemical Physics.

Deep learning helps make predictions on the effects of the El Niño-Southern Oscillation on river flows more accurate.

Machine learning models help identify important environmental properties that influence how often extreme rain events occur with critical intensity and duration.

PNNL researchers combined deep operator networks with spectral methods to efficiently and accurately solve complex equations.

Hydrogen bonding and proton transfer control the identity and reactivity of molecules in highly concentrated and alkaline aluminate solutions.

Data-driven autonomous technology to rapidly design and deliver antiviral interventions targeting SARS-CoV-2 to reduce drug discovery timeline and advance bio preparedness capabilities.

The work by the team at PNNL takes a critical step in leveraging ML to accelerate advanced manufacturing R&D, specifically for manufacturing techniques without access to efficient, first-principles simulations.