DOE Report Envisions Acceleration of Research With Help From Artificial Intelligence

Staff from Department of Energy national laboratories released over the summer a report that describes an initiative that seeks to produce artificial intelligence tools to accelerate research and development efforts relevant to the agency’s mission.

The report, titled “Advanced Research Directions on AI for Science, Energy, and Security,” envisions the application of AI in multiple areas, including the development of foundation models dedicated to addressing scientific and national security problems, the use of AI to operate machinery and experiments, and the use of AI to determine the properties of a given material, the American Institute of Physics website reported.

According to Rick Stevens, head of Argonne National Laboratory’s Computing, Environment and Life Sciences Directorate, the DOE’s experience in handling multidisciplinary teams under the Exascale Computing Project, puts it in a position to take the lead in AI R&D.

“This is not going to be a small initiative. This could be several times or more the scale of what we have been doing with ECP,” Stevens told the Advanced Scientific Computing Advisory Committee in June.

The AI initiative was conceived as a follow-on to the ECP, an effort led by the DOE and the National Nuclear Security Administration that seeks to establish an ecosystem of exascale computing solutions to provide high levels of computational power for use in various kinds of research.

Oak Ridge National Laboratory already launched its exascale supercomputer — dubbed Frontier — last year. Argonne National Laboratory installed the final components of its exascale supercomputer — dubbed Aurora — this June. Lawrence Livermore National Laboratory began installing components of its exascale supercomputer — dubbed El Capitan — this July.