Tue 6 Nov 2018 08:50 - 09:15 at Cabot - AI SEPS Chair(s): Yukinori Sato, Ali Jannesari

In this paper we present PIRA – an infrastructure for automatic instrumentation refinement for performance analysis. It automates the process to generate an initial performance overview measurement and gradually refines it, based on the recorded runtime information. This relieves a performance analyst from the time consuming and largely manual, yet mechanical, task of selecting which functions to capture in subsequent measurements. PIRA implements an existing aggregation strategy that heuristically determines which functions to include or exclude for initial overview measurements. Moreover, it implements a newly developed heuristic to incorporate profile information and expand instrumentation in hot-spot regions only. The approach is evaluated on different benchmarks, including the SU2 multi-physics solver package. PIRA is able to generate instrumentation configurations that contain the application’s hot-spot, but generate significantly less overhead when compared to the Score-P reference measurement.

Tue 6 Nov
Times are displayed in time zone: (GMT-05:00) Guadalajara, Mexico City, Monterrey change

08:00 - 10:00: AI-SEPS - AI SEPS at Cabot
Chair(s): Yukinori SatoToyohashi University of Technology, Ali JannesariIowa State University
seps-2018-papers08:00 - 08:50
PrabhatNERSC, Berkeley Lab
seps-2018-papers08:50 - 09:15
Jan-Patrick LehrGraduate School of Computational Engineering, TU Darmstadt, Alexander HückInstitute for Scientific Computing, TU Darmstadt, Christian BischofScientific Computing, TU Darmstadt
seps-2018-papers09:15 - 09:30
Yohann UguenUniv Lyon, INSA Lyon, Inria, CITI, Eric PetitIntel, France
seps-2018-papers09:30 - 10:00
Yukinori SatoToyohashi University of Technology, Ali JannesariIowa State University, Shigeru ChibaThe University of Tokyo