A GPU-based survey for millisecond radio transients using ARTEMIS
Institute for the Future of Computing, Oxford Martin School, Oxford
arXiv:1111.6399v1 [astro-ph.IM] (28 Nov 2011)
@article{2011arXiv1111.6399A,
author={Armour, W. and Karastergiou, A. and Giles, M. and Williams, C. and Magro, A. and Zagkouris, K. and Roberts, S. and Salvini, S. and Dulwich, F. and Mort, B.},
title={"{A GPU-based survey for millisecond radio transients using ARTEMIS}"},
journal={ArXiv e-prints},
archivePrefix={"arXiv"},
eprint={1111.6399},
primaryClass={"astro-ph.IM"},
keywords={Instrumentation and Methods for Astrophysics},
year={2011},
month={nov}
}
Astrophysical radio transients are excellent probes of extreme physical processes originating from compact sources within our Galaxy and beyond. Radio frequency signals emitted from these objects provide a means to study the intervening medium through which they travel. Next generation radio telescopes are designed to explore the vast unexplored parameter space of high time resolution astronomy, but require High Performance Computing (HPC) solutions to process the enormous volumes of data that are produced by these telescopes. We have developed a combined software /hardware solution (code named ARTEMIS) for real-time searches for millisecond radio transients, which uses GPU technology to remove interstellar dispersion and detect millisecond radio bursts from astronomical sources in real-time. Here we present an introduction to ARTEMIS. We give a brief overview of the software pipeline, then focus specifically on the intricacies of performing incoherent de-dispersion. We present results from two brute-force algorithms. The first is a GPU based algorithm, designed to exploit the L1 cache of the NVIDIA Fermi GPU. Our second algorithm is CPU based and exploits the new AVX units in Intel Sandy Bridge CPUs.
November 29, 2011 by hgpu
