Undergraduate Research

Research Overview

The following work was done as part of the NSF funded Research Experience for Undergraduate (REU) program at the Cyclotron Institute at Texas A&M University.

Chiral effective field theory is an approximation to quantum chromodynamics, which provides a modern framework for understanding the structure and dynamics of nuclear many-body systems. Sophisticated nuclear two- and three-body forces are used to calculate the equation of state in many-body perturbation theory. Recent works have had much success in applying the theory to describe the ground- and excited-state properties of light and medium-mass atomic nuclei when combined with ab initio numerical techniques. This research aims to extend the application of chiral effective field theory to describe the nuclear equation of state required for supercomputer simulations of core-collapse supernovae. Given the large range of densities, temperatures, and proton fractions probed during stellar core collapse, microscopic calculations of the equation of state require large computational resources on the order of one million CPU hours. Graphics processing units (GPUs) are investigated to significantly reduce the computational cost of these calculations, which will enable a more accurate and precise description of this important input to numerical astrophysical simulations.

The results of the parallelized program (using GPUs) vs. the serial program (CPU) are summarized in the figure below. The plot shows the time it takes a program to run on CPU vs. GPU as a function of mesh points. These mesh points could be thought of as precision points. The more precise the final needs to be, the more mesh points are needed. Additionally, the higher precision points require more computational time. The GPU code performs much better than the CPU code at any given number of mesh points, and especially higher numbers of mesh points. This is partially due to the fact that GPU can accommodate to higher memory demand than CPU can.

Selected Presenations: