(Image credit: Argonne National Laboratory/US Department of Energy)
Our understanding of the Universe has taken a major step forward after the Frontier supercomputer at Oak Ridge National Laboratory (ORNL) created a simulation of the Universe on a previously unachievable scale.
Frontier used software called Hardware/Hybrid Accelerated Cosmology Code (HACC) as part of the ExaSky project, which was part of the US Department of Energy's (DOE) $1.8 billion Exascale Computing Project, the largest software research and development initiative supported by the DOE.
Exasky’s science applications were designed to run 50 times faster than previous standards, but Frontier and HACC far exceeded expectations, running nearly 300 times faster than similar simulations of Saturn’s moon Titan. The DoE/HACC team has spent the seven years since the first simulation improving the capabilities of exascale supercomputers like Frontier.
This allowed for hydrodynamic cosmological simulations, which are much more computationally sophisticated models that take into account principles such as the expansion of the universe and the influence of dark matter. Previously used models included only measurements of gravity, gas, or plasma.
The power of exascale computing
The simulation, conducted in November 2024, involved approximately 9,000 Frontier compute nodes, all equipped with AMD Instinct MI250X graphics cards.
Frontier is the second-fastest supercomputer in the world, with a processing power of up to 1.4 exaflops. Supercomputer performance is measured in floating-point operations per second (FLOPS) — where one floating-point operation represents a mathematical calculation.
Anything that delivers more than 999 petaflops (0.9 exaflops) of performance is called an “exascale” supercomputer. The only other machine with more power than Frontier is El Capitan, which can reach 1.7 exaflops.
In addition to modeling the universe, Frontier has been used in other important studies. In April 2023, scientists created the Cloud-Resolving Simple Atmosphere Model E3SM (SCREAM), a program that simulated a full year of global climate data at a resolution of just over 3 km. One of the most sophisticated climate models ever computed, it is now a cornerstone in analyzing the complex interactions between the atmosphere, oceans, and land, leading to better weather forecasts and more accurate climate change data.
In the field of materials, Frontier has enabled researchers to create new substrates and geometries for materials with improved properties, making them stronger, lighter, and more resistant to corrosion. Its exascale computing capabilities have enabled researchers to model chemical interactions at the molecular level to predict the behavior of materials. The supercomputer has also played a key role in the search for new materials for energy storage, transportation, manufacturing, and nuclear medicine.
However, scientists are particularly interested in how exascale computing can enrich artificial intelligence (AI). The high speed of these machines allows programmers to quickly test algorithms and analyze large amounts of data, especially in tasks such as creating faster large language models or applying supercomputers to climate models and forecasting climate change.
Drew TurneySo
Sourse: www.livescience.com
