Benchmark Repository

This repository contains detailed information on a growing number of benchmark programs. The information was obtained with tools such as program analyzers, timers, profilers, and hardware monitors.
Some information characterizes aspects of a program, independent of the machine it runs on.
(Example: the call structure of a program)
Subroutine/Loop Call Graph (Ursa Minor tool)
Other information is machine dependent.
(Example: timing results)
SPECseis with small dataset executed on a variety of machines.

Information Categories

The repository includes a number of information categories, such as application characterization, measured performance, and advanced model analysis. To be as consistet as possible, all categories are listed in all benchmarks. However, not all categories are populated. Also, more categories may be added in the future. (For a list and brief explanation of the information categories Click here ).

Benchmarks

Currently data is available for the following benchmark programs:
 
SPECseis
a seismic processing application, part of the SPEChpc suite.		 SPECchem
a chemistry simulation of molecules, part of the SPEChpc suite.		 SPECclimate
a mesoscale weather model, part of the SPEChpc suite.
Image of the word PURDUE spelled out using point diffusers in a seismic simulation and migration applicaiton. Thymine molecule, "medium" dataset of SPECchem Storm over Korean Pennensula, "medium" dataset of SPECclimate

Contributions to this Repository

Both data contributions and suggestions on how this service can be made more useful to the community are welcome. Currently there is no format requirement to provide new data. We will try to adapt your format. Though, in the future this may need to be changed. 
R. Eigenmann

Last modified: Mon Nov 15 13:45:41 PST 1999