Personal tools
You are here: Home Workshops Summer 2012 Performance Tools for Extreme-scale Computing
Document Actions

Performance Tools for Extreme-scale Computing

by John Mellor-Crummey last modified 2012-06-30 19:58

June 26-29, 2012, Snowbird Ski & Summer Resort, Snowbird, Utah, USA

Untitled Document



Multipetaflop systems for open science are being installed at Argonne National Laboratory (ANL) and Oak Ridge National Laboratory (ORNL). 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. For ANL's Mira system, the principal challenge will be exploiting threading effectively on each node's 16-core, 4-way multithreaded processors. For ORNL's Titan system, the principal challenge will be exploiting hybrid parallelism across a 16-core processor and a GPU.

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 at scale will require tools with unprecedented capabilities.

The goal of this workshop is to bring together tools researchers to discuss challenges of performance analysis and debugging 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 tools infrastructure, debugging, performance instrumentation, measurement, analysis, and visualization. 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 emerging extreme scale systems.
  • 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.

Final Program

Tuesday, June 26, 2012
  7:30 AM Breakfast Buffet (Magpie B Room)
Tue AM 1 9:00 AM The Deconstruction of Dyninst in 2012 Bill Williams
Drew Bernat
University of Wisconsin
9:30 AM HPCToolkit for Heterogeneous Architectures Milind Chabbi
Rice University
10:00 AM Profiling OpenMP with HPCtoolkit Xu Liu
Rice University
  10:30 AM Morning Break
Tue AM 2 11:00 AM CBTF and its Integration into the Structure of Open|SpeedShop Jim Galarowicz
Krell Labs
11:30 AM Optimization of Floating-point Precision using Binary Modification Michael Lam
University of Maryland
12:00 PM TAUdb: Refactoring the TAU PerfDMF Kevin Huck
University of Oregon
  12:30 PM Lunch (Magpie B Room)
Tue PM 1 2:00 PM The Marriage of MRNET and GPUs Ben Welton
University of Wisconsin
2:30 PM nVidia Tools for Profiling and Monitoring David Goodwin
3:00 PM BSC Tools Update: Using clustering and Folding Judit Gimenez
Barcelona Supercomputer Center
  3:30 PM Afternoon Break
Tue PM 2 4:00 PM GPGPU efforts at LANL, Status and Tool Needs, plus CBTF for Monitoring Dave Montoya
Los Alamos National Lab
4:30 PM A Lightweight OpenMP Runtime Alexandre Eichenberger
5:00 PM New Tools and Challenges for BlueGene/Q Dong Ahn, Matt Legendre
Lawrence Livermore National Lab
  7:00 PM Dinner (Atrium Lounge)
Wednesday, June 27, 2012
  7:30 AM Breakfast Buffet (Magpie B Room)
Wed AM 1 9:00 AM Linux Kernel and Performance Tool Support for AMD PMUs Robert Richter
9:30 AM An Update on perf_events Stefane Eranian
Google, Inc.
10:00 AM The Upcoming PAPI 5.0 Release Vince Weaver
Univeristy of Tennessee
  10:30 AM Morning Break
Wed AM 2 11:00 AM Tool Support for Hybrid Systems Heidi Poxon
11:30 AM Performance Visualization of an AMR Load Balancer (and more) Martin Schulz,
Lawrence Livermore National Lab
12:00 PM MIAMI: Machine-independent Recipes for Performance Insight Gabriel Marin
Oak Ridge National Lab
  12:30 PM Lunch (Magpie B Room)
Wed PM 1 2:00 PM A Software Development Environment for Earth System Modeling Depei Qian
Beihang University
2:30 PM Task Analysis with Score-P Daniel Lorenz
Juelich Supercomputer Center
3:00 PM Node-wide Performance (and Power) Introspection Applied to Scheduling Rob Fowler
  3:30 PM Afternoon Break      
Wed PM 2 4:00 PM The Gooda Tool
(Generic Cycle Accounting)
(Driving the Visualizer)
David Levinthal
Google, Inc.
4:30 PM The MPI_T Interface and Its Implementation in Mpich2 Dave Goodell
Argonne National Lab
5:00 PM LIBI for Optimal Tool and Application Start-Up Dorian Arnold
University of New Mexico
  7:00 PM Dinner (Atrium Lounge)
Working Group Outbrief
OpenMP Tools API Report (preliminary notes to be expanded)
Hardware counter working group
GPU performance tool working group
Tools at Scale
Thursday, June 28, 2012
  7:30 AM Breakfast Buffet (Magpie B Room)
Thu AM 1 9:00 AM Working Groups - Planning and Initial Meetings
  10:30 AM Morning Break
Thu AM 2 11:00 AM Working Groups Meet
  12:30 PM Lunch (Magpie B Room)
Thu PM 2:00 PM Working Groups Meet
  7:00 PM Dinner (Atrium Lounge)
Friday, June 29, 2012
  7:30 AM Breakfast Buffet (Magpie B Room)
Fri AM 1 9:00 AM Working Groups Meet
  10:30 AM Morning Break
Fri AM 2 11:00 AM Working Group Out-Briefs
  12:30 PM Lunch (Magpie B Room)
Fri PM 1 2:00 PM Follow-up Discussions
  3:30 PM Afternoon Break      
Fri PM 2 4:00 PM Follow-up Discussions
  7:00 PM Dinner (Atrium Lounge)
Saturday, June 30, 2012
« March 2017 »
Su Mo Tu We Th Fr Sa

Powered by Plone

CScADS Collaborators include:

Rice University ANL UCB UTK WISC