Implementation of the Neuberger-Dirac operator on GPUs
Institut fur Kernphysik, Johannes Gutenberg-Universitat Mainz, Johann Joachim Becher-Weg 45, 55099 Mainz, Germany
arXiv:1010.5636 [hep-lat] (27 Oct 2010)
@article{walk2010implementation,
title={Implementation of the Neuberger-Dirac operator on GPUs},
author={Walk, B. and Wittig, H. and Dranischnikow, E. and Sch{\”o}mer, E.},
journal={Arxiv preprint arXiv:1010.5636},
year={2010}
}
Recent developments have shown that a lot can be gained for QCD simulations from GPU hardware. This can be exploited especially in the case of Ginsparg-Wilson fermions when the com putational costs are particularly high. In this work, we use the Neuberger-Dirac operator as our realisation of Ginsparg-Wilson fermions, which greatly facilitate lattice investigations of decays like $K to pipi$. We report on the ongoing study of our GPU implementation of the Neuberger-Dirac operator including the exact treatment of the low lying eigenmodes of the Wilson-Dirac operator. Our benchmarks show that we achieve speed-up factors of around 23 and 16 in single and double precision, respectively.
November 9, 2010 by hgpu