We give an overview of QPACE 2, which is a custom-designed supercomputer based on Intel Xeon Phi processors, developed in a collaboration of Regensburg University and Eurotech. We give some general recommendations for how to write high-performance code for the Xeon Phi and then discuss our implementation of a domain-decomposition-based solver and present a number of benchmarks.
Cite
@article{arxiv.1502.04025,
title = {QPACE 2 and Domain Decomposition on the Intel Xeon Phi},
author = {Paul Arts and Jacques Bloch and Peter Georg and Benjamin Glaessle and Simon Heybrock and Yu Komatsubara and Robert Lohmayer and Simon Mages and Bernhard Mendl and Nils Meyer and Alessio Parcianello and Dirk Pleiter and Florian Rappl and Mauro Rossi and Stefan Solbrig and Giampietro Tecchiolli and Tilo Wettig and Gianpaolo Zanier},
journal= {arXiv preprint arXiv:1502.04025},
year = {2015}
}
Comments
plenary talk at Lattice 2014, to appear in the conference proceedings PoS(LATTICE2014), 15 pages, 9 figures