Adaptive algebraic multigrid on SIMD architectures
Department of Physics, University of Regensburg, 93040 Regensburg, Germany
arXiv:1512.04506 [physics.comp-ph], (14 Dec 2015)
@article{heybrock2015adaptive,
title={Adaptive algebraic multigrid on SIMD architectures},
author={Heybrock, Simon and Rottmann, Matthias and Georg, Peter and Wettig, Tilo},
year={2015},
month={dec},
archivePrefix={"arXiv"},
primaryClass={physics.comp-ph}
}
We present details of our implementation of the Wuppertal adaptive algebraic multigrid code DD-alpha AMG on SIMD architectures, with particular emphasis on the Intel Xeon Phi processor (KNC) used in QPACE 2. As a smoother, the algorithm uses a domain-decomposition-based solver code previously developed for the KNC in Regensburg. We optimized the remaining parts of the multigrid code and conclude that it is a very good target for SIMD architectures. Some of the remaining bottlenecks can be eliminated by vectorizing over multiple test vectors in the setup, which is discussed in the contribution of Daniel Richtmann.
December 15, 2015 by hgpu
