Personal tools
You are here: Home Workshops Summer 2009 Workshop on Performance Tools for Petascale Computing
Document Actions

Workshop on Performance Tools for Petascale Computing

by admin last modified 2009-07-23 11:38

July 20-23, 2009, Granlibakken Resort and Conference Center, Tahoe City, CA, USA

Organizers



Abstract

Petascale computer systems are now available to the DOE science community; a contract to install a 20 petaflop machine at Lawrence Livermore National Laboratory in 2011 has just been announced. Compared to systems of the past, these machines represent a dramatic increase in scale and complexity. Making effective use of systems at this scale will require performance tools that help application developers by pinpointing and explaining application performance bottlenecks. Two performance challenges for applications are to make effective use of processor nodes, and to keep parallelization overhead low enough so that extreme-scale parallelism is practical.

Providing tools that make it possible to address these problems is requiring increasingly sophisticated methods for instrumentation, measurement, analysis and modeling of application performance. Meeting these challenges will require tools with unprecedented capabilities.

Today, research and development in the area of performance tools for parallel systems is fragmented. Individual groups have developed tools and components with unique capabilities. Working together to capitalize these capabilities into shared community infrastructure would reduce duplication of effort, accelerate development of effective tools with increasing sophistication, and simplify the process of developing tools for new systems as they emerge.

The goal of this workshop is to bring together performance tools researchers to discuss challenges of performance analysis on emerging petascale systems, review ongoing research, and work together to tackle the challenges of performance analysis on petascale systems together as a community. Attendees at the workshop will be tools developers in the areas of performance instrumentation, measurement, analysis, and modeling. We expect that about a third of the workshop will consist of opportunities to work together in smaller groups to make progress on tackling problems of mutual interest.

Specific aims of the workshop include

  • Identifying performance tool capabilities needed to analyze the spectrum of issues that arise with petascale platforms.
  • Discussing emerging capabilities of research and commercial tools in this space.
  • Identifying common needs, functionality, and opportunities for sharing infrastructure.
  • Discussing design aspects of sharable components.
  • Developing standardized interfaces to facilitate sharing.


Agenda

Sunday, July 19, 2009
7:00 PM Dinner Buffet  
 
 
Monday, July 20, 2009
8:00 AM Breakfast  
9:00 AM The Latest and Greatest in the Dyninst Binary Code Toolkit
  Matt Legendre, Bill Williams, Madhavi Krishnan, and Drew Bernat,
University of Wisconsin
 
10:00 AM New Developments around Scalasca
  Felix Wolf,
Juelich Supercomputing Centre
 
10:30 AM Morning Break
11:00 AM Memory Subsystem Profiling with the Sun Studio Performance Analyzer (with demo)
  Marty Itzkowitz, Sun Microsystems
 
11:45 AM Linux Has a Generic Performance Monitoring API! (with demo)
  Stephane Eranian, Google
 
12:30 PM Lunch
2:00 PM Combining static and dynamic analysis for debugging and performance tuning
Abstract   Slides
  Martin Schulz, LLNL
 
2:30 PM Performance Measurement and Analysis of Multithreaded Programs
  Nathan Tallent, Rice University
 
3:00 PM Afternoon Break
3:30 PM What It Takes To Assign Blame
  Jeff Hollingsworth and Nick Rutar,
University of Maryland
 
4:00 PM Performance Strategies for Parallel Mathematical Libraries Based on Historical Knowledgebase
  Eduardo Cesar and Ania Morajko,
Autonomous University of Barcleona
 
4:30 PM A Proposal for a Profiling Data Exchange Format
  Bernd Mohr, JSC
Martin Schulz, LLNL
Dan Gunter, LBL
Kevin Huck, University Oregon
Xingfu Wu, Texas A&M
 
6:30 PM Dinner
 
 
Tuesday, July 21, 2009
8:00 AM Breakfast
9:00 AM HPCToolkit Update 2009
  Mike Fagan, Rice University
 
10:00 AM Performance Measurement and Analysis of Heterogeneous Systems: Tasks and GPU Accelerators
  Allen Malony, University Oregon
 
10:30 AM Morning Break
11:00 AM Automatic Profiling Analysis
  Luiz Derose and Heidi Poxon, Cray
 
11:30 AM Obtaining Extremely Detailed Information at Scale
  Jesus Labarta, BSC
 
12:00 AM Building a Community Infrastructure for Scalable On-Line Performance Analysis Tools
  Jim Galarowicz, Krell Institute
David Montoya, LANL
 
12:30 PM Lunch
2:00 PM Autonomous Tool Infrastructure
  Dorian Arnold, University of New Mexico
 
2:30 PM Scalable Tool Infrastructure for the Cray XT Using Tree-Based Overlay Networks
  Philip Roth, ORNL
 
3:00 PM Afternoon Break
3:30 PM ROSE Open Compiler Infrastructure supporting Custom Tools for Software Analysis, Transformation, and Optimization
Abstract   Slides
  Dan Quinlan, LLNL
 
4:00 PM Performance Tuning using PMU Features of Core i7
  Ramesh Peri, Intel
 
4:30 PM Semi-Automatic Models of Communication Volume and Frequency for SPMD Applications
  Gabriel Marin, ORNL
 
6:30 PM Dinner
 
 
Working Group Materials
Instruction Sampling and Profiling
Marty Itzkowitz, Sun Microsystems
Infrastructure Working Group
Bronis de Supinski et al.
 
 
 
Wednesday, July 22, 2009
8:00 AM Breakfast
9:00 AM Working Groups - Planning and initial meetings
10:30 AM Morning Break
11:00 AM Working Groups
12:30 PM Lunch
2:00 PM Informal discussions of research challenges, ongoing tools research, working group activities, and collaboration plans
6:30 PM Dinner
 
 
Thursday, July 23, 2009
8:00 AM Breakfast
9:00 AM Working Groups
10:30 AM Morning Break
11:00 AM Working Groups
12:30 PM Lunch
2:00 PM Working Groups
3:00 PM Afternoon Break
3:30 PM Working Groups and Closing Discussion
6:30 PM Dinner

Sponsors

This workshop is sponsored by the Center for Scalable Application Development Software, with funding from the DOE Office of Science through the Scientific Discovery through Advanced Computing (SciDAC) program.

« November 2017 »
Su Mo Tu We Th Fr Sa
1234
567891011
12131415161718
19202122232425
2627282930
 

Powered by Plone

CScADS Collaborators include:

Rice University ANL UCB UTK WISC