Folding Methods for Event Timelines in Performance Analysis

The complexity of today's high performance computing systems, and their parallel software, requires performance analysis tools to fully understand application performance behavior. The visualization of event streams has proven to be a powerful approach for the detection of various types of performance problems. However, visualization of large numbers of process streams quickly hits the limits of available screen resolution. To alleviate this problem we propose folding strategies for event timelines that consider common questions during performance analysis. We demonstrate the effectiveness of our solution with code inefficiencies in two real-world applications, PIConGPU and COSMO-SPECS. Our methods facilitate visual scalability and provide powerful overviews of performance data at the same time. Furthermore, our folding strategies improve GPU stream visualization and allow easy evaluation of the GPU device utilization.

[1]  Holger Brunst,et al.  Custom Hot Spot Analysis of HPC Software with the Vampir Performance Tool Suite , 2012, Parallel Tools Workshop.

[2]  Jesús Labarta,et al.  New Analysis Techniques in the CEPBA-Tools Environment , 2009, Parallel Tools Workshop.

[3]  Juan Gonzalez,et al.  Automatic detection of parallel applications computation phases , 2009, 2009 IEEE International Symposium on Parallel & Distributed Processing.

[4]  Martin Schulz,et al.  Alignment-Based Metrics for Trace Comparison , 2013, Euro-Par.

[5]  Michael Wagner,et al.  Hierarchical Memory Buffering Techniques for an In-Memory Event Tracing Extension to the Open Trace Format 2 , 2013, 2013 42nd International Conference on Parallel Processing.

[6]  Jack J. Dongarra,et al.  An algebra for cross-experiment performance analysis , 2004, International Conference on Parallel Processing, 2004. ICPP 2004..

[7]  William Gropp,et al.  Toward Scalable Performance Visualization with Jumpshot , 1999, Int. J. High Perform. Comput. Appl..

[8]  Wolfgang E. Nagel,et al.  Performance Optimization for Large Scale Computing: The Scalable VAMPIR Approach , 2001, International Conference on Computational Science.

[9]  Martin Schulz,et al.  Open | SpeedShop: An open source infrastructure for parallel performance analysis , 2008, Sci. Program..

[10]  Lucas Mello Schnorr,et al.  A hierarchical aggregation model to achieve visualization scalability in the analysis of parallel applications , 2012, Parallel Comput..

[11]  Bernd Mohr,et al.  Usage of the SCALASCA toolset for scalable performance analysis of large-scale parallel applications , 2008, Parallel Tools Workshop.

[12]  Martin Schulz,et al.  Stack Trace Analysis for Large Scale Debugging , 2007, 2007 IEEE International Parallel and Distributed Processing Symposium.

[13]  H Burau,et al.  PIConGPU: A Fully Relativistic Particle-in-Cell Code for a GPU Cluster , 2010, IEEE Transactions on Plasma Science.

[14]  Karen L. Karavanic,et al.  Scalable Event Trace Visualization , 2009, Euro-Par Workshops.

[15]  B.P. Miller,et al.  MRNet: A Software-Based Multicast/Reduction Network for Scalable Tools , 2003, ACM/IEEE SC 2003 Conference (SC'03).

[16]  Holger Brunst Integrative concepts for scalable distributed performance analysis and visualization of parallel programs , 2008 .

[17]  Barton P. Miller What to Draw? When to Draw? An Essay on Parallel Program Visualization , 1993, J. Parallel Distributed Comput..

[18]  V. Grützun,et al.  Simulation of the influence of aerosol particle characteristics on clouds and precipitation with LM-SPECS: Model description and first results , 2008 .