Project Cycle Time Dramatically Reduced! Scientists Use Supercomputing To Fuel Laser Fusion Research

Recently, the University of Rochester's Laboratory for Laser Energetics (LLE) installed a new supercomputer to support its laser fusion experiments.
The new supercomputer increases the lab's computing power by a factor of four and reduces the time required to complete certain projects from 30 weeks to a few days.
The University of Rochester's Laboratory for Laser Energetics (LLE) is one of only a handful of facilities in the world that study laser-driven inertial confinement fusion (ICF), which scientists use for national security purposes and to derive energy from nuclear fusion.
"A new supercomputer located at the university will allow researchers to simulate complex, high-energy-density phenomena in ICF in three dimensions with unprecedented detail," said Valeri Goncharov, director of the laboratory's theoretical division and a scientist.
"For example, it is very difficult, if not impossible, to directly measure the evolution of micrometer-scale target defects in an implosion. However, detailed 3D simulations can model how such phenomena alter more easily measurable experimental observations," Goncharov explains, "Discovering the correlation between simulation results and experimental data will help to determine the importance of sub-scale target features and other complex physical effects in experiments. "
The machine, called Conesus, was built by Intel and developed in collaboration with Dell Technologies and Lawrence Livermore National Laboratory (LLNL). It is one of only seven Intel fourth-generation Sapphire Rapids systems in the world and one of only two in the United States.
The "World's Top 500 Supercomputers" (TOP500 List) program, started in 1993, publishes an updated list of the world's most powerful supercomputers twice a year.
How will laser fusion experiments benefit?
The University of Rochester's Laboratory for Laser Energetics has two very powerful lasers - the Omega and Omega EP - that researchers use to conduct studies, including those involving ICF. This work builds on a breakthrough in ignition, the fusion reaction that produces a net energy gain, that scientists made last year at LLNL's National Ignition Facility (NIF).
William Scullin, head of the lab's high-performance computing group, said, "About 10 times a day, our lasers are used to create an energetic star in a jar."
But the path to laser-driven inertial confinement fusion (ICF) begins with supercomputers modeling the materials, the lasers, and the experiment itself.
Scullin said, "We have 1D, 2D and 3D modeling capabilities to simulate inertial confinement fusion. We model materials and plasmas at extreme temperatures and pressures. High-power lasers are not commercially available components. Therefore, we design many of our own optical and laser systems in-house. In addition, there is an increasing amount of statistical work to be done."
According to Scullin, as the need for statistical analysis increases, computational scientists are exploring how to use machine learning to find out what's going on from old and new data. To make these discoveries possible, LLE needs new computing resources.
Conesus will provide computational resources for scientists to collect more data and conduct higher-resolution studies, including using machine learning on larger data sets, according to Scullin. Projects that might have taken 30 weeks to complete on earlier systems can be completed in days using Conesus.
Conesus has several projects planned, including testing statistical models of the cryogenic implosion of the Omega laser system; simulating alpha particle stopping and burning plasma; and studying liquid crystals, which produce large responses with very high thermal stability.
The University of Rochester's Laboratory for Laser Energetics (LLE) will house two 25 gigawatt lasers as part of a project supported by the National Science Foundation (NSF) at the University of Rochester that has a budget of $18 million over a three-year period. As part of the project, the lab will establish a new facility called EP-OPAL, which will be dedicated to the study of the interaction of ultrahigh-intensity lasers with matter.