TY  - CONF
AU  - Vierjahn, Tom
AU  - Hermanns, Marc-André
AU  - Mohr, Bernd
AU  - Müller, Matthias S.
AU  - Kuhlen, Torsten W.
AU  - Hentschel, Bernd
TI  - Using Directed Variance to Identify Meaningful Views in Call-path Performance Profiles
CY  - Piscataway, NJ, USA
PB  - IEEE Press
M1  - FZJ-2016-06939
SN  - 978-1-5090-5226-4
SP  - 9-16
PY  - 2016
AB  - Understanding the performance behaviour of massively parallel high-performance computing (HPC) applications based on call-path performance profiles is a time-consuming task. In this paper, we introduce the concept of directed variance in order to help analysts find performance bottlenecks in massive performance data and in the end optimize the application. According to HPC experts' requirements, our technique automatically detects severe parts in the data that expose large variation in an application's performance behaviour across system resources. Previously known variations are effectively filtered out. Analysts are thus guided through a reduced search space towards regions of interest for detailed examination in a 3D visualization. We demonstrate the effectiveness of our approach using performance data of common benchmark codes as well as from actively developed production codes.
T2  - The 3rd International Workshop on Visual Performance Analysis
CY  - 18 Nov 2016 - 18 Nov 2016, Salt Lake City, Utah (USA)
Y2  - 18 Nov 2016 - 18 Nov 2016
M2  - Salt Lake City, Utah, USA
LB  - PUB:(DE-HGF)8 ; PUB:(DE-HGF)7
DO  - DOI:10.1109/vpa.2016.7
UR  - https://juser.fz-juelich.de/record/824330
ER  -