Preparation and optimization of a diverse workload for a large-scale heterogeneous system

Autor: Martin Schulz, Ulrike Meier Yang, David F. Richards, Tong Chen, Shiv Sundram, Todd Gamblin, Shelby Lockhart, Phil Regier, David Beckingsale, Ed Zywicz, Ruipeng Li, Giacomo Domeniconi, James C. Sexton, Bob Walkup, Jarom Nelson, Carlos Costa, Hui-Fang Wen, Ramesh Pankajakshan, John A. Gunnels, Xiaohua Zhang, Brian Van Essen, Kathryn M. O'Brien, I-Feng W. Kuo, Johann Dahm, Guillaume Thomas-Collignon, Bert Still, Naoya Maruyama, Jamie A. Bramwell, David Boehme, Kathleen Shoga, Carol S. Woodward, Howard A. Scott, M. P. Katz, Ian Karlin, T Epperly, Tzanio V. Kolev, Eun Kyung Lee, Steven H. Langer, Christopher Ward, David J. Gardner, Sara I. L. Kokkila-Schumacher, Christopher Young, Kevin O'Brien, Barry Chen, Björn Sjögreen, Jose R. Brunheroto, Claudia Misale, Roger Pearce, Guojing Cong, Matthew Legendre, Lu Wang, Jaime H. Moreno, Kathleen McCandless, Cyril Zeller, Rao Nimmakayala, Bronis R. de Supinski, Xinyu Que, Sorin Bastea, Robert D. Falgout, Peng Wang, Charway R. Cooper, Aaron Fisher, Jim Brase, R. Neely, David Appelhans, Alexey Voronin, James N. Glosli, Slaven Peles, Pei-Hung Lin, Tony Degroot, Hai Le, Daniel A. White, Levi Barnes, Steve Rennich, Yoonho Park, Peter D. Barnes, Bob Anderson, Jonathan J. Wong, Robert C. Blake

Publikováno v: SC

Productivity from day one on supercomputers that leverage new technologies requires significant preparation. An institution that procures a novel system architecture often lacks sufficient institutional knowledge and skills to prepare for it. Thus, t

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_________::4e88bb7115e2e0999663b946908ca92b
https://doi.org/10.1145/3295500.3356192

Poster: Memory and Parallelism Exploration Using the LULESH Proxy Application

Autor: Jeff Keasler, Ian Karlin, Bert Still, Edgar A. León, Esthela Gallardo, Jim McGraw

Publikováno v: SC Companion

Current and planned computer systems present challenges for scientific programming. Memory capacity and bandwidth are limiting performance as floating point capability increases due to more cores per processor and wider vector units. Effectively usin

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_________::2bc5b0d6b0cdd3d5b0ce94aee541fac8
https://doi.org/10.1109/sc.companion.2012.235