weiqunzhang/asterx

AsterX is a GPU-accelerated GRMHD code for dynamical spacetimes

AsterX is a GPU-accelerated GRMHD code for dynamical spacetimes, written in C++. It is built upon the CarpetX driver, which is intended for the Einstein Toolkit. CarpetX is based on AMReX, a software framework for block-structured AMR (adaptive mesh refinement).

Full documentation will soon be available at asterx.readthedocs.io.

• 

Overview

• Heavily derived from the GRMHD code Spritz.
• Solves the GRMHD equations in 3D Cartesian coordinates and on dynamical spacetimes using high-resolution shock capturing (HRSC) schemes.
• Based on the flux-conservative Valencia formulation.
• Directly evolves the staggered vector potential.

Available modules

• AsterX - the core GRMHD module
• AsterSeeds - initial data module
• Con2PrimFactory - module providing different conservative-to-primitive variable recovery routines
• EOSX - equation of state driver
• ReconX - provider of different reconstruction schemes
• TOVSolver - a modified version of the publicly available TOVSolver thorn used within the Einstein Toolkit

Getting started

Instructions for downloading and building the Einstein Toolkit including CarpetX can be found here.

Details for building and running AsterX along with CarpetX will be added to asterx.readthedocs.io soon..

Related talks and tutorials

• "Using CarpetX: A Guide for Early Adopters". Recorded seminar talk by Erik Schnetter, providing an overview of the current capabilities of CarpetX.
• "Tutorial: GPUs and the Einstein Toolkit". Recorded tutorial by Lorenzo Ennoggi, Jay Kalinani and Federico Lopez Armengol during the North American Einstein Toolkit workshop 2022, presenting a brief overview on AsterX, followed by a hands-on session.
• "AsterX: a new open-source GPU-accelerated GRMHD code for dynamical spacetimes". Slides based on the talk by Jay Kalinani at the APS April Meeting 2023.