Scientists Use Supercomputer To Unravel Mysteries of Dark Matter and the Universe’s Evolution

Physics Astronomy Quantum Mechanics
Nikola

 The largest-ever simulations of the Lyman-𝛼 forest spectral data carried out by the PRIYA supercomputer show the large-scale structure of the universe.

The brightest light in the cosmos is produced by distant quasars, which shine like cosmic beacons. Light emission from these quasars surpasses that of our Milky Way galaxy as well. This enormous light comes from the disintegration of stuff under the intense pressure of a supermassive black hole. Astronomers rely on cosmological parameters as vital numerical instruments because they let them study the universe’s development billions of years after the Big Bang.


Advancements in Simulation Technology

The Frontera supercomputer at the Texas Advanced Computing Center (TACC), supported by the National Science Foundation (NSF), assisted astronomers in creating PRIYA, the broadest suite of hydrodynamic simulations to date for recreating large-scale structure in the universe, using quasar light data.

Simeon Bird, an assistant professor of astronomy at the University of California, Riverside, stated, “We’ve created a new simulation model to compare data that exists in the real universe.”

PRIYA was built by Bird et al. using optical light data from the Sloan Digital Sky Survey (SDSS) Extended Baryon Oscillation Spectroscopic Survey (eBOSS). In October 2023, he and associates announced PRIYA in the Journal of Cosmology and Astroparticle Physics (JCAP).

“We juxtapose the eBOSS data against an array of simulation models that have disparate cosmological parameters and initial conditions to the cosmos, including disparate matter densities,” said Bird. The optimum option is identified, together with the maximum deviation from it that won’t compromise the reasonable agreement between the simulations and the data. This information reveals the amount of matter and/or structure present in the universe.

The Role of PRIYA in Cosmological Research

The PRIYA simulation suite is linked to ASTRID, a large-scale cosmological modeling program that Bird co-developed and is used to research early universe re-ionization, the coalescence of supermassive black holes, and galaxy formation. PRIYA goes above and above. It modifies the beginning conditions using the black hole production laws and galaxy information from ASTRID.

“By using these rules, we can modify the initial conditions of the model we developed to fit galaxies and black holes, and then compare it with the Lyman-𝛼 forest data from eBOSS of the neutral hydrogen gas,” explained Bird.


The “forest” of densely clustered absorption lines on a graph of the quasar spectrum, which results from electron transitions between energy levels in neutral hydrogen atoms, is what gives rise to the moniker “Lyman-𝛼 forest.” The location, mass, and temperature of massive interstellar neutral hydrogen clouds are shown by the “forest.” Furthermore, the lumpiness of the gas suggests the existence of dark matter, a theorized component that pulls on galaxies while not being visible.

Refining Cosmological Parameters with PRIYA

Coauthors Simeon Bird and M.A. Fernandez and Ming-Feng Ho of UC Riverside, along with their work published to JCAP in September 2023, employed PRIYA simulations to fine-tune cosmological parameters.
Prior studies of the neutrino mass parameters disagreed with observations made of the Cosmic Microwave Background radiation (CMB), which is thought to be the remnants of the Big Bang explosion. Tight limits on the mass of neutrinos are set by astronomers using CMB data from the Plank Space Observatory. Determining the mass of neutrinos is crucial for cosmological models that depict the large-scale structure of the universe since they are the most prevalent particle in the cosmos.
Using much larger and more intricately structured simulations than ever before, we performed a new analysis. After that, the differences with the Planck CMB data vanished and were replaced by a new tension that resembled other large-scale structure measurements at low redshifts, according to Bird. “The study’s primary finding is the confirmation of the σ8 tension between weak lensing and CMB measurements up to redshift 2, or ten billion years ago.”


σ8, or the number of neutral hydrogen gas structures on a size of 8 megaparsecs, or 2.6 million light years, is one well-constrained parameter from the PRIYA study. This shows how many dark matter clumps are present in that area, according to Bird.
The scalar spectral index, or ns, was another parameter that was limited. It has to do with how the clumsiness of dark matter changes when one increases the size of the region under study. It shows how quickly the universe expanded in the first few minutes following the Big Bang.
The initial behavior of the cosmos is determined by the scalar spectral index. Determining the universe’s beginning conditions and the behavior of its high-energy physics is the main goal of PRIYA, according to Bird.

The Impact of Supercomputing on Cosmological Studies

Bird clarified that the PRIYA simulations required supercomputers due to their large size.
According to Bird, “PRIYA simulations require so much memory that they can only be run on supercomputers.”
A Leadership Resource Allocation on the Frontera supercomputer was given to Bird by TACC. The UC Riverside High-Performance Computer Cluster’s resources were also used for analysis computations.
Using more than 100,000 core hours, the PRIYA simulations on Frontera are among the largest cosmological simulations ever performed. They simulate a system of 3072^3 (or around 29 billion) particles in a “box” that is 120 megaparsecs on edge, or roughly 3.91 million light-years wide. On Frontera, PRIYA simulations used more than 600,000 node hours.


Frontera was crucial to the research because we needed to run a lot of these simulations on a supercomputer large enough to perform one of them reasonably easily. We could not fix them without something like Frontera. They just wouldn’t be able to run at all, therefore it wouldn’t take very long, according to Bird.

PRIYA simulation data was also stored for a long time using TACC’s Ranch system.

“Ranch is significant because it allows us to utilize PRIYA for more projects. Bird stated, “This could double or triple our impact in science.”

Bird said, “Our appetite for more computer power is insatiable.” “It’s incredible that we are observing the majority of the universe while sitting here on this small planet.”

Related Posts

Scroll top