hpcviewer:
Interactive Presentation of Performance

The HPCToolkit Performance Tools

2012/09/21

Version 5.3.2-r4407

The Java-based hpcviewer interactively presents program performance in a top-down fashion.

Table of Contents

Synopsis

Command-line usage:
      hpcviewer [options] [hpctoolkit-database]

GUI usage:
      Launch hpcviewer and open the Experiment database [hpctoolkit-database].

Description

The Java-based hpcviewer interactively presents program-performance Experiment databases in a top-down fashion. Since Experiment databases are self-contained, they may be relocated from a cluster for visualization on a laptop or workstation.

Arguments

hpctoolkit-database
An HPCToolkit Experiment database, which is the result of executing hpcprof, hpcprof-mpi or hpcprof-flat.

Options

-n
Do not display the Callers View. (Saves memory and time.)

-consolelog
Send log entries to a console in addition to a log file.

-debug
Log additional information about plug-in dependency problems.

Detailed Description

Views

hpcviewer supports three principal views of an application's performance data. Both inclusive (costs of a procedure including all its callees) and exclusive costs (costs excluding callees) are reported.

Panes

The browser window is divided into three panes:

Thread-Centric Plots

hpcviewer can display plot graphs of thread-level metric values. This is useful for quickly assessing load imbalance across processes and threads.

To create a graph, use the calling context view and select a scope by left-clicking a line in the navigation pane. Then, right-click the selected scope to show the context menu. (This menu begins with `Zoom-in' and `Zoom-out.') At the bottom of the context menu is a list of metrics that hpcviewer can graph. Each metric contains a sub-menu that lists the three different types of graphs that hpcviewer can plot. The Plot graph sorts the processes by process and thread number. The Sorted plot graph sorts the processes by metric value. And the Histogram graph shows a bar graph of the frequency of metric values.

Note: currently, these plot graphs are available only with databases created by hpcprof-mpi and not by hpcprof. See the section on ``Plotting Graphs of Thread-level Metric Values'' in the User's Manual for more description and a view of a plot graph.

See Also

hpctoolkit(1) .

Version

Version: 5.3.2-r4407 of 2012/09/21.

License and Copyright

Copyright
© 2002-2013, Rice University.
License
See README.License.

Authors

Laksono Adhianto
John Mellor-Crummey
Nathan Tallent
Rice HPCToolkit Research Group
Email: hpctoolkit-forum =at= rice.edu
WWW: http://hpctoolkit.org.