Scalable critical-path based performance analysis

David Böhme, Felix Wolf, Bronis R. De Supinski, Martin Schulz, Markus Geimer

Publikation: Beitrag in Buch/Bericht/KonferenzbandKonferenzbeitragBegutachtung

46 Zitate (Scopus)

Abstract

The critical path, which describes the longest execution sequence without wait states in a parallel program, identifies the activities that determine the overall program runtime. Combining knowledge of the critical path with traditional parallel profiles, we have defined a set of compact performance indicators that help answer a variety of important performance-analysis questions, such as identifying load imbalance, quantifying the impact of imbalance on runtime, and characterizing resource consumption. By replaying event traces in parallel, we can calculate these performance indicators in a highly scalable way, making them a suitable analysis instrument for massively parallel programs with thousands of processes. Case studies with real-world parallel applications confirm that - in comparison to traditional profiles - our indicators provide enhanced insight into program behavior, especially when evaluating partitioning schemes of MPMD programs.

OriginalspracheEnglisch
TitelProceedings of the 2012 IEEE 26th International Parallel and Distributed Processing Symposium, IPDPS 2012
Seiten1330-1340
Seitenumfang11
DOIs
PublikationsstatusVeröffentlicht - 2012
Extern publiziertJa
Veranstaltung2012 IEEE 26th International Parallel and Distributed Processing Symposium, IPDPS 2012 - Shanghai, China
Dauer: 21 Mai 201225 Mai 2012

Publikationsreihe

NameProceedings of the 2012 IEEE 26th International Parallel and Distributed Processing Symposium, IPDPS 2012

Konferenz

Konferenz2012 IEEE 26th International Parallel and Distributed Processing Symposium, IPDPS 2012
Land/GebietChina
OrtShanghai
Zeitraum21/05/1225/05/12

Fingerprint

Untersuchen Sie die Forschungsthemen von „Scalable critical-path based performance analysis“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren