Grid-based minimization at scale: Feldman-Cousins corrections for SBN

Holger Schulz; Marianette Wospakrik; Mark Ross-Lonergan; Guanqun Ge; Saba Sehrish; Marc Paterno; Jim Kowalkowski; Wes Ketchum; Georgia Karagiorgi

Grid-based minimization at scale: Feldman-Cousins corrections for SBN

Computational Physics 2020-02-20 v1 High Energy Physics - Experiment

Authors: Holger Schulz , Marianette Wospakrik , Mark Ross-Lonergan , Guanqun Ge , Saba Sehrish , Marc Paterno , Jim Kowalkowski , Wes Ketchum , Georgia Karagiorgi

View on arXiv ↗ PDF ↗

Abstract

We present a computational model for the construction of Feldman-Cousins (FC) corrections frequently used in High Energy Physics (HEP) analysis. The program contains a grid-based minimization and is written in C++. Our algorithms exploit vectorization through Eigen3, yielding a single-core speed-up of 350 compared to the original implementation, and achieve MPI data parallelism by using DIY. We demonstrate the application to scale very well at High Performance Computing (HPC) sites. We use HDF5 in conjunction with HighFive to write results of the calculation to file.

Keywords

graphics processing unit computing computational physics

Cite

@article{arxiv.2002.07858,
  title  = {Grid-based minimization at scale: Feldman-Cousins corrections for SBN},
  author = {Holger Schulz and Marianette Wospakrik and Mark Ross-Lonergan and Guanqun Ge and Saba Sehrish and Marc Paterno and Jim Kowalkowski and Wes Ketchum and Georgia Karagiorgi},
  journal= {arXiv preprint arXiv:2002.07858},
  year   = {2020}
}

Grid-based minimization at scale: Feldman-Cousins corrections for SBN

Abstract

Keywords

Cite

Related papers