Cactus Framework: Black Holes to Gamma Ray Bursts

TL;DR

This paper discusses the development of the Cactus computational framework for simulating gamma-ray bursts, emphasizing scalability to petaflop supercomputers and addressing future architectural challenges.

Contribution

It introduces a petascale approach and discusses the scalability and future development plans of the Cactus framework for astrophysical simulations.

Findings

Cactus framework supports petascale astrophysical simulations.

Scalability challenges are identified for future supercomputing architectures.

Plans for enhancing Cactus to meet petascale computing demands.

Abstract

Gamma Ray Bursts (GRBs) are intense narrowly-beamed flashes of gamma-rays of cosmological origin. They are among the most scientifically interesting astrophysical systems, and the riddle concerning their central engines and emission mechanisms is one of the most complex and challenging problems of astrophysics today. In this article we outline our petascale approach to the GRB problem and discuss the computational toolkits and numerical codes that are currently in use and that will be scaled up to run on emerging petaflop scale computing platforms in the near future. Petascale computing will require additional ingredients over conventional parallelism. We consider some of the challenges which will be caused by future petascale architectures, and discuss our plans for the future development of the Cactus framework and its applications to meet these challenges in order to profit from these new architectures.