Large-Scale High-Lundquist Number Reduced MHD Simulations of the Solar Corona Using GPU Accelerated Machines

TL;DR

This paper demonstrates the acceleration of large-scale reduced MHD simulations of the solar corona using GPU hardware, significantly improving computational efficiency for high Lundquist number regimes.

Contribution

It introduces a GPU-accelerated implementation of a 3D RMHD code for coronal heating studies, enabling longer simulations at higher resolutions than previously possible.

Findings

GPU acceleration improves simulation speed significantly.

Performance varies across different hardware configurations.

The method enables more detailed coronal heating modeling.

Abstract

We have recently carried out a computational campaign to investigate a model of coronal heating in three-dimensions using reduced magnetohydrodynamics (RMHD). Our code is built on a conventional scheme using the pseudo-spectral method, and is parallelized using MPI. The current investigation requires very long time integrations using high Lundquist numbers, where the formation of very fine current layers challenge the resolutions achievable even on massively parallel machines. We present here results of a port to Nvidia CUDA (Compute Unified Device Architecture) for hardware acceleration using graphics processing units (GPUs). In addition to a brief discussion of our general strategy, we will report code performance on several machines which span a variety of hardware configurations and capabilities. These include a desktop workstation with commodity hardware, a dedicated research workstation equipped with four Nvidia C2050 GPUs, as well as several large-scale GPU accelerated distributed memory machines: Lincoln/NCSA, Dirac/NERSC, and Keeneland/NICS.