Robert Bell, Allen D. Malony, Sameer Shende Department of Computer and Information Science Computational Science.

Slides:



Advertisements
Similar presentations
Machine Learning-based Autotuning with TAU and Active Harmony Nicholas Chaimov University of Oregon Paradyn Week 2013 April 29, 2013.
Advertisements

K T A U Kernel Tuning and Analysis Utilities Department of Computer and Information Science Performance Research Laboratory University of Oregon.
Dynamic performance measurement control Dynamic event grouping Multiple configurable counters Selective instrumentation Application-Level Performance Access.
Sameer Shende Department of Computer and Information Science Neuro Informatics Center University of Oregon Tool Interoperability.
Allen D. Malony, Sameer Shende Department of Computer and Information Science Computational Science Institute University.
Profiling S3D on Cray XT3 using TAU Sameer Shende
TAU Parallel Performance System DOD UGC 2004 Tutorial Allen D. Malony, Sameer Shende, Robert Bell Univesity of Oregon.
The TAU Performance Technology for Complex Parallel Systems (Performance Analysis Bring Your Own Code Workshop, NRL Washington D.C.) Sameer Shende, Allen.
Nick Trebon, Alan Morris, Jaideep Ray, Sameer Shende, Allen Malony {ntrebon, amorris, Department of.
On the Integration and Use of OpenMP Performance Tools in the SPEC OMP2001 Benchmarks Bernd Mohr 1, Allen D. Malony 2, Rudi Eigenmann 3 1 Forschungszentrum.
Allen D. Malony, Sameer Shende Department of Computer and Information Science Computational Science Institute University.
The TAU Performance System: Advances in Performance Mapping Sameer Shende University of Oregon.
Performance Instrumentation and Measurement for Terascale Systems Jack Dongarra, Shirley Moore, Philip Mucci University of Tennessee Sameer Shende, and.
Allen D. Malony Department of Computer and Information Science Computational Science Institute University of Oregon TAU Performance.
June 2, 2003ICCS Performance Instrumentation and Measurement for Terascale Systems Jack Dongarra, Shirley Moore, Philip Mucci University of Tennessee.
Instrumentation and Profiling David Kaeli Department of Electrical and Computer Engineering Northeastern University Boston, MA
TAU: Performance Regression Testing Harness for FLASH Sameer Shende
Allen D. Malony, Sameer Shende Department of Computer and Information Science Computational Science Institute University.
Allen D. Malony, Sameer Shende, Robert Bell Department of Computer and Information Science Computational Science Institute, NeuroInformatics.
Kai Li, Allen D. Malony, Robert Bell, Sameer Shende Department of Computer and Information Science Computational.
Sameer Shende, Allen D. Malony Computer & Information Science Department Computational Science Institute University of Oregon.
Performance Technology for Complex Parallel Systems REFERENCES.
Understanding and Managing WebSphere V5
1 Babak Behzad, Yan Liu 1,2,4, Eric Shook 1,2, Michael P. Finn 5, David M. Mattli 5 and Shaowen Wang 1,2,3,4 Babak Behzad 1,3, Yan Liu 1,2,4, Eric Shook.
1 Performance Analysis with Vampir DKRZ Tutorial – 7 August, Hamburg Matthias Weber, Frank Winkler, Andreas Knüpfer ZIH, Technische Universität.
Lecture 8. Profiling - for Performance Analysis - Prof. Taeweon Suh Computer Science Education Korea University COM503 Parallel Computer Architecture &
SC’13: Hands-on Practical Hybrid Parallel Application Performance Engineering Introduction to VI-HPS Brian Wylie Jülich Supercomputing Centre.
Computer and Automation Research Institute Hungarian Academy of Sciences Presentation and Analysis of Grid Performance Data Norbert Podhorszki and Peter.
Integrated Performance Views in Charm++: Projections meets TAU Scott Biersdorff Allen D. Malony Department Computer and Information Science University.
Scalable Analysis of Distributed Workflow Traces Daniel K. Gunter and Brian Tierney Distributed Systems Department Lawrence Berkeley National Laboratory.
John Mellor-Crummey Robert Fowler Nathan Tallent Gabriel Marin Department of Computer Science, Rice University Los Alamos Computer Science Institute HPCToolkit.
Technology + Process SDCI HPC Improvement: High-Productivity Performance Engineering (Tools, Methods, Training) for NSF HPC Applications Rick Kufrin *,
Profile Analysis with ParaProf Sameer Shende Performance Reseaerch Lab, University of Oregon
Martin Schulz Center for Applied Scientific Computing Lawrence Livermore National Laboratory Lawrence Livermore National Laboratory, P. O. Box 808, Livermore,
1 Performance Analysis with Vampir ZIH, Technische Universität Dresden.
Visualization Workshop David Bock Visualization Research Programmer National Center for Supercomputing Applications - NCSA University of Illinois at Urbana-Champaign.
Performance Analysis Tool List Hans Sherburne Adam Leko HCS Research Laboratory University of Florida.
Service - Oriented Middleware for Distributed Data Mining on the Grid ,劉妘鑏 Antonio C., Domenico T., and Paolo T. Journal of Parallel and Distributed.
Profiling, Tracing, Debugging and Monitoring Frameworks Sathish Vadhiyar Courtesy: Dr. Shirley Moore (University of Tennessee)
Dynamic performance measurement control Dynamic event grouping Multiple configurable counters Selective instrumentation Application-Level Performance Access.
Debugging parallel programs. Breakpoint debugging Probably the most widely familiar method of debugging programs is breakpoint debugging. In this method,
Workshop BigSim Large Parallel Machine Simulation Presented by Eric Bohm PPL Charm Workshop 2004.
PerfExplorer Component for Performance Data Analysis Kevin Huck – University of Oregon Boyana Norris – Argonne National Lab Li Li – Argonne National Lab.
GEON2 and OpenEarth Framework (OEF) Bradley Wallet School of Geology and Geophysics, University of Oklahoma
Allen D. Malony Department of Computer and Information Science TAU Performance Research Laboratory University of Oregon Discussion:
High-Performance and Grid Computing for Neuroinformatics: NIC and Cerebral Data Systems Allen D. Malony University of Oregon Professor Department of Computer.
Allen D. Malony, Sameer S. Shende, Robert Bell Kai Li, Li Li, Kevin Huck Department of Computer.
Scientific Programmes Committee Centre for Aerospace Systems Design & Engineering Amitay Isaacs Department of Aerospace Engineering Indian Institute of.
Shangkar Mayanglambam, Allen D. Malony, Matthew J. Sottile Computer and Information Science Department Performance.
Integrated Performance Views in Charm++: Projections meets TAU Scott Biersdorff Allen D. Malony Department Computer and Information Science University.
Parallel Performance Measurement of Heterogeneous Parallel Systems with GPUs Allen D. Malony, Scott Biersdorff, Sameer Shende, Heike Jagode†, Stanimire.
Performane Analyzer Performance Analysis and Visualization of Large-Scale Uintah Simulations Kai Li, Allen D. Malony, Sameer Shende, Robert Bell Performance.
Other Tools HPC Code Development Tools July 29, 2010 Sue Kelly Sandia is a multiprogram laboratory operated by Sandia Corporation, a.
CEPBA-Tools experiences with MRNet and Dyninst Judit Gimenez, German Llort, Harald Servat
Online Performance Analysis and Visualization of Large-Scale Parallel Applications Kai Li, Allen D. Malony, Sameer Shende, Robert Bell Performance Research.
PERFORMANCE OF THE OPENMP AND MPI IMPLEMENTATIONS ON ULTRASPARC SYSTEM Abstract Programmers and developers interested in utilizing parallel programming.
Parallel OpenFOAM CFD Performance Studies Student: Adi Farshteindiker Advisors: Dr. Guy Tel-Zur,Prof. Shlomi Dolev The Department of Computer Science Faculty.
Performance Tool Integration in Programming Environments for GPU Acceleration: Experiences with TAU and HMPP Allen D. Malony1,2, Shangkar Mayanglambam1.
Kai Li, Allen D. Malony, Sameer Shende, Robert Bell
Productive Performance Tools for Heterogeneous Parallel Computing
Performance Technology for Scalable Parallel Systems
Tracing and Performance Analysis Tools for Heterogeneous Multicore System by Soon Thean Siew.
TAU integration with Score-P
Allen D. Malony, Sameer Shende
TAU Parallel Performance System
TAU: A Framework for Parallel Performance Analysis
Allen D. Malony Computer & Information Science Department
Outline Introduction Motivation for performance mapping SEAA model
Parallel Program Analysis Framework for the DOE ACTS Toolkit
TAU Performance DataBase Framework (PerfDBF)
Presentation transcript:

Robert Bell, Allen D. Malony, Sameer Shende Department of Computer and Information Science Computational Science Institute / NeuroInformatics Center University of Oregon ParaProf : A Portable, Extensible, and Scalable Tool for Parallel Performance Profile Analysis

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool Outline  Motivation  ParaProf Objectives  Related Work  ParaProf Features and Functionality  Examples  512-processor SAMRAI execution  Interactive demonstration  Software engineering of ParaProf  Recent advancements  Future work  Concluding remarks

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool Motivation  Profiling is well-known and broadly applied technique  Profiling tools are not the same  Different profile instrumentation and measurement  Sequential vs. parallel profiling  System-specific, proprietary, and incompatible  Complicates cross-platform performance studies  Slows development of portable, robust profile analysis  Increased detail and complexity of profile data  Hardware performance counters  Integration of system and application performance data  Parallel profile data / analysis and large-scale parallelism

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool ParaProf Objectives  Portable, extensible, and scalable tool for profile analysis  Offer “best of breed” capabilities to performance analysts  Build as profile analysis framework for extensibility  Work with different (most) types of profile data  Support input of profile data from different sources  Universal performance profile analysis capabilities  Large-scale analysis and display support  Multi-profile (multi-experiment)  Programmable analysis  Modular, object-oriented software engineering  Broadly applied

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool Related Work  Rich history of sequential and parallel profiling tools  Sequential profilers  prof and gprof  Unix profiling of execution time using sampling method  gprof includes callgraph profiling (parent-child distribution)  cxperf and ssrun (SGI)  Hardware performance counter profiling  vprof (Visual Profiler)  DynaProf  PAPI-based profiling using dynamic instrumentation  HPCView  Support for multiple profile analysis

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool Related Work (continued)  Parallel profilers  GuideView and VGV  OpenMP applications (VGV also supports MPI profiling)  Proprietary  Aksum  Targeted to Linux systems with multiple experiment support  SvPablo  Cross-platform with source-based views  Expert  Trace-generated profile data  Performance property/problem analysis and display  HPM Toolkit

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool ParaProf Features  Parallel profile data  “Experiment” gives profile for every thread of execution  Multiple performance metrics (time, HPC, …)  Based on TAU performance system  Event-based profiles  Support for callpath profiles  Profile data input  Post-mortem from raw files  Post-mortem from performance database  Online from running program (in progress)  Multiple experiment profiles active simultaneously

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool ParaProf Features (continued)  Profile analysis  Statistical analysis per thread and across threads  Individual events and event groups  Value-based and percent-based analysis  Derived statistics and distribution statistics for scalability  Experiment profile integration  Profile performance displays  Bargraph displays  Hyperlink navigation

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool TAU Performance System Framework  Tuning and Analysis Utilities (aka Tools Are Us)  Performance system framework for scalable parallel and distributed high-performance computing  Targets a general complex system computation model  nodes / contexts / threads  Multi-level: system / software / parallelism  Measurement and analysis abstraction  Integrated toolkit for performance instrumentation, measurement, analysis, and visualization  Portable performance profiling/tracing facility  Open software approach

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool TAU Performance System Architecture EPILOG Paraver

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool ParaProf Architecture

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool ParaProfile Manager

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool ParaProf Profile Display (VTF)

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool Full Profile Display (SAMRAI) 512 processes

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool Profile Statistics Histogram (SAMRAI)  Need to address profile display scalability  Statistical analysis to show performance distributions  Value histogramming showing # threads in value range  Define # bins and value distribution function Execution time (wallclock)Floating point operations

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool Recent ParaProf Enhancements  Integration of ParaProf with DynaProf  Convert DynaProf profile data to TAU format

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool Future Work  Profile translators  Sequential: prof/grof (vprof), cxperf/ssrun  Parallel: SvPablo, Aksum, HPM Toolkit  Cross-experiment analysis  Generalized programmable analysis engine  Integration with online performance profiling in TAU  Online profile monitor in TAU currently available  Analysis of profiles generated from trace phase analysis  Trace-based phase profile tool in development  More sophisticated performance display graphics  Use 3D performance visualization library (in progress)

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool Concluding Remarks  ParaProf is a portable parallel profile analysis tool  ParaProf provides broad, integrated functionality  Designed to analyze and display large-scale profile  Designed for multi-experiment performance studies  Intended to serve as a universal profile analysis system  Robust design and software engineering  Future work on extended analysis and visualization  Future work on performance database integration

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool More Information  TAU performance system  Acknowledgements  DOE project, “Performance Technology for Tera-Class Parallel Computers: Evolution of the TAU Performance System,”

EuroPar, August :53ParaProf: A Portable, Extensible Profile Analysis Tool