15 Jul 2024

 

Science Breakthrough

Using simulations performed on NERSC’s Perlmutter system, a team of researchers used lattice quantum chromodynamics (lattice QCD) – a computational approach to obtaining quantities in hadronic physics by discretizing space-time on a four-dimensional lattice – to understand for the first time certain aspects of proton structure in terms of its fundamental quark and gluon constituents. Their research was recently published in Physical Review Letters.

 

Science Background

For decades, scientists have known that protons and other hadrons have a substructure of quarks and gluons, and were able to understand the distribution of charge within the proton using electromagnetic form factors. However, the distribution of other physical properties of the proton, like its mass, its spin, its internal forces, and how they are split between quarks and gluons, have been less well understood. All of that information is encoded in the gravitational form factors of hadrons, which have been challenging to measure.

In recent years, some of the gravitational form factors of the proton have been experimentally measured for the first time; in this paper, the researchers gained a better understanding of how energy, angular momentum, pressure, and shear forces are distributed inside the proton by calculating the full set of its gravitational form factors for the first time. In doing so, using lattice QCD, they showed that the gluons inside the proton are more spread out than the quarks, corresponding to a larger “gluonic” than quark radius.

This new information answers long-standing essential questions about the substructure of protons as well as pointing researchers toward more targeted questions about other hadrons.

 

Science Breakdown

The researchers performed their calculations on NERSC’s Perlmutter system using Chroma and QUDA, parts of a suite of software produced by NVIDIA and the USQCD collaboration of U.S. scientists developing HPC tools for lattice QCD. The project, “The gluonic structure, pressure, and shear distributions of the nucleon,” (PI: Shanahan) used 552,000 GPU hours on Perlmutter.

Because the data for measuring the contributions of the quarks and gluons was both very complex and very noisy, the team generated a statistical sample of over four million measurements. Each measurement was computationally intensive, and was made feasible by the power of Perlmutter’s GPUs.

 

[Image]

For the first time, researchers have a better understanding of how gluons (left side) and quarks (right side) form the substructure of protons and other hadrons.