Fahringer Thomas ; Rico-Gallego Juan A. ; Grasso Ivan ; Hristov Atanas ; Lastovetsky Alexey L. ; Karatza Helen ; Fabrizio Marozzo ; Dana Petcu ; Stavrinides Georgios L. ; Domenico Talia ; Astsatryan Hrachya ; Trunfio Paolo
Extreme scale parallel computing systems will have tens of thousands of optionally accelerator-equipped nodes with hundreds of cores each, as well as deep memory hierarchies and complex interconnect topologies. Such exascale systems will provide hardware parallelism at multiple levels and will be energy constrained. Their extreme scale and the rapidly deteriorating reliability of their hardware components means that exascale systems will exhibit low mean-time-between failure values. Furthermore, existing programming models already require heroic programming and optimization efforts to achieve high efficiency on current supercomputers. Invariably, these efforts are platform-specific and non-portable. In this article, we explore the shortcomings of existing programming models and runtimes for large-scale computing systems. We propose and discuss important features of programming paradigms and runtimes to deal with exascale computing systems with a special focus on data-intensive applications and resilience. Finally, we discuss code sustainability issues and propose several software metrics that are of paramount importance for code development for ultrascale computing systems.
oai:noad.sci.am:136121
Supercomputing Frontiers and Innovations
Institute for Informatics and Automation Problems ; Institut de Recherche en Informatique de Toulouse ; Universidad de Extremadura ; University for Information Science and Technology "St. Paul the Apostle"Ohrid ; University College Dublin ; West University of Timisoara ; Università della Calabria ; University of Innsbruck ; Aristotle University of Thessaloniki
Armenia ; France ; Spain ; North Macedonia ; Romania ; Italy ; Austria ; Greece ; Ireland
Apr 19, 2021
Apr 19, 2021
153
https://noad.sci.am/publication/149479
Հրատարակության անուն | Ամսաթիվ |
---|---|
Georges Da Costa, Exascale Machines Require New Programming Paradigms and Runtimes | Apr 19, 2021 |