Poster: Correlating Sub-Phenomena in Performance Data in the Frequency Domain

Tom Vierjahn, Marc-André Hermanns, Bernd Mohr, Matthias Stefan Müller, Torsten Wolfgang Kuhlen, Bernd Hentschel
LDAV 2016 – The 6th IEEE Symposium on Large Data Analysis and Visualization

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 separately. We provide the mathematical framework and an overview over the visualization, and we demonstrate the effectiveness of our technique.

Best Poster Award!

» Show BibTeX
@inproceedings{Vierjahn-2016-03, Author = {Vierjahn, Tom and Hermanns, Marc-Andr\'{e} and Mohr, Bernd and M\"{u}ller, Matthias S. and Kuhlen, Torsten W. and Hentschel, Bernd}, Booktitle = {LDAV 2016 -- Posters (accepted)}, Date-Added = {2016-08-31 22:14:47 +0000}, Date-Modified = {2016-08-31 22:15:58 +0000}, Title = {Correlating Sub-Phenomena in Performance Data in the Frequency Domain} }



Disclaimer Home Visual Computing institute RWTH Aachen University