The Department of Energy’s (DOE) Argonne National Laboratory has launched a new artificial intelligence (AI) inference service designed to help researchers accelerate scientific discovery by providing cloud-like access to advanced AI models running on Argonne’s high-performance computing systems.
The newly launched ALCF Inference Service is powered by Argonne Leadership Computing Facility (ALCF) systems. It gives researchers access to large language models, science foundation models, and computer vision models for analyzing large datasets and testing new scientific ideas.
The service is intended to speed the path from data collection to scientific insight.
“Our inference service helps close the gap between developing AI models and putting them to work in scientific research,” said Michael Papka, director of the ALCF.
“By offering AI inference as a shared resource, we enable researchers to apply AI at scale to their data, simulations, and experiments, without the burden of building and maintaining their own infrastructure,” Papka said.
Inference refers to the use of trained AI models to analyze information, identify patterns, and make predictions. While consumer AI tools use inference to answer questions in real time, Argonne officials said the same technology can help scientists guide experiments, interpret complex data, and perform analytical tasks more efficiently.
“Inference services allow researchers to spend less time managing models and more time testing hypotheses,” said Venkat Vishwanath, AI and machine learning lead at the ALCF.
“Instead of taking days or weeks to analyze data, scientists can rapidly interpret results, refine experiments, and explore complex systems in ways that weren’t practical before,” the official said.
The service is viewed as a key enabler for the Department of Energy’s Genesis Mission, a national AI initiative aimed at building what officials describe as the world’s most powerful scientific platform to accelerate discovery science, strengthen national security, and drive energy innovation.
The inference service also will support the American Science Cloud, the Genesis Mission’s integrated environment linking DOE supercomputers, experimental facilities, and data resources.
Argonne said the service provides researchers access to models, including Google’s Gemma series, Meta’s LLaMA models, and OpenAI’s GPT-OSS family, along with Argonne-developed systems such as AuroraGPT. Unlike vendor-managed frontier AI models that may change without notice, the ALCF service is designed to provide a more stable and transparent environment for scientific research, the lab said.
The service is already being used by researchers across the DOE national laboratory system, including scientists at Brookhaven National Laboratory, Fermi National Accelerator Laboratory, Los Alamos National Laboratory, Lawrence Berkeley National Laboratory, Lawrence Livermore National Laboratory, Oak Ridge National Laboratory, Sandia National Laboratories, and Thomas Jefferson National Accelerator Facility. Officials said the growing adoption supports integrated, multi-institutional research workflows.
Argonne said the inference service is expected to advance research across numerous scientific fields. In fusion energy research, AI models can analyze experimental data in real time and predict plasma disruptions before they occur. In high-energy physics and astronomy, the service can help researchers process massive amounts of collider and telescope data to identify rare events and new phenomena more quickly.
The inference service also supports chemistry and materials science research through projects such as ChemGraph, an AI framework that automates molecular simulation workflows.
“This allows scientists to explore more candidate molecules, iterate on designs faster and manage large-scale calculations as an integrated process rather than a series of disconnected jobs,” said Murat Keçeli, an Argonne computational scientist who helped develop ChemGraph.