headerlogo
scyourway

SC Conference - Activity Details



HPCToolkit: Performance Tools for Parallel Scientific Computing

Authors:
Laksono Adhianto  (Rice University)
Sinchan Banerjee  (Massachusetts Institute of Technology)
Michael Fagan  (Rice University)
Mark Krentel  (Rice University)
Gabriel Marin  (Rice University)
John Mellor-Crummey  (Rice University)
Nathan Tallent  (Rice University)
Posters Session
Tuesday,  05:15PM - 07:00PM
Room Rotunda Lobby
Abstract:
HPCToolkit is an integrated suite of tools that supports measurement, analysis, attribution, and presentation of application performance for both sequential and parallel programs. HPCToolkit supports pinpointing and quantifying scalability bottlenecks in parallel programs with a measurement overhead of only a few percent. In 2008, new capabilities were added to HPCToolkit for collecting call path profiles for fully optimized codes on systems with x86_64 and ppc64 processors, pinpointing and quantifying bottlenecks in multithreaded programs, exploring performance information and source code using a new user interface, and displaying hierarchical space-time diagrams based on traces of asynchronous call stack samples. In fall 2008, HPCToolkit will be deployed on the TACC 500TF Ranger system. After kernel support for statistical sampling using hardware performance counters is in place, HPCToolkit will be deployed on Cray XT and Blue Gene systems at Department of Energy Laboratories.
   IEEE Computer Society  /  ACM     2 0   Y E A R S   -   U N L E A S H I N G   T H E   P O W E R   O F   H P C