TY - CONF
AU - Vierjahn, Tom
AU - Hermanns, Marc-Andre
AU - Mohr, Bernd
AU - Muller, Matthias S.
AU - Kuhlen, Torsten W.
AU - Hentschel, Bernd
TI - Correlating sub-phenomena in performance data in the frequency domain
PB - IEEE
M1 - FZJ-2018-00859
SN - 978-1-5090-5659-0
SP - 105-106
PY - 2016
AB - Finding and understanding correlated performance behaviour of the individual functions of massively parallel high-performance computing (HPC) applications is a time-consuming task. In this poster, we propose filtered correlation analysis for automatically locating interdependencies in call-path performance profiles. Transforming the data into the frequency domain splits a performance phenomenon into sub-phenomena to be correlated
T2 - 2016 IEEE 6th Symposium on Large Data Analysis and Visualization
CY - 23 Oct 2016 - 28 Oct 2016, Baltimore, MD (USA)
Y2 - 23 Oct 2016 - 28 Oct 2016
M2 - Baltimore, MD, USA
LB - PUB:(DE-HGF)8 ; PUB:(DE-HGF)7
DO - DOI:10.1109/LDAV.2016.7874340
UR - https://juser.fz-juelich.de/record/842652
ER -