The READEX formalism for automatic tuning for energy efficiency

• DOI: 10.1007/s00607-016-0532-7
• Publication Date: 2017-01-10T13:46:09Z
• Authors: Joseph Schuchart; Michael Gerndt; Per Gunnar Kjelds...; Michael Lysaght; David Horák; Lubomír Říha; Andreas Gocht; Mohammed Sourouri; Madhura Kumaraswamy; Anamika Chowdhury; Magnus Jahre; Kai Diethelm; Othman Bouizi; Umbreen Sabir Mian; Jakub Kružík; Radim Sojka; Martin Beseda; Venkatesh Kannan; Zakaria Bendifallah; Daniel Hackenberg; Wolfgang E. Nagel
• Keywords: Exascale computing; Dynamism; Formalism (music)

Energy efficiency is an important aspect of future exascale systems, mainly due to rising energy cost. Although High performance computing (HPC) applications are compute centric, they still exhibit varying computational characteristics in different regions the program, such as compute-, memory-, and I/O-bound code regions. Some today's clusters already offer mechanisms adjust system resource requirements application, e.g., by controlling CPU frequency. However, manually tuning for improved a tedious painstaking task that often neglected application developers. The European Union's Horizon 2020 project READEX (Runtime Exploitation Application Dynamism Energy-efficient eXascale computing) aims at developing tools-aided approach current HPC applications. To reach this goal, combines technologies from two ends spectrum, embedded systems HPC, constituting split design-time/runtime methodology. From domain, Periscope Tuning Framework (PTF) extended perform dynamic auto-tuning fine-grained using scenario methodology, which was originally developed improving systems. This paper introduces concepts project, its envisioned implementation, preliminary results demonstrate feasibility approach.