.
Providing a New Generation of Economically Viable
High Productivity Computing Systems
.

Scalable Synthetic Compact Application (SSCA)
Benchmarks


    • SSCA #1 - Bioinformatics Optimal Pattern Matching; stresses integer and character operations (no floating point required); compute-limited and embarrassingly parallel. Written and Matlab V7.0 SP2 Executable Specifications.
    • SSCA #2 - Graph Analysis; stresses memory access; uses integer and character operations (no floating point required); compute-intensive and hard to parallelize.
    • SSCA #3 - SAR Sensor Processing, Knowledge Formation, and File IO; computationally taxing: Principal performance goal is rate at which answers are generated. Significant file IO component; if a time-optimal data storage scheme is picked, it may not be optimal for data retrieval; large block data transfers, stressful memory access patterns, and small file IO.
      • (v1.0a download) (unofficial) C code by Meng-Ju of UMD and Janice Onanian McMahon of USC/ISI
      • I/O-only version of SSCA#3 C code by Bradley Kuszmaul of MIT
      • (v1.03 download) (official version) Written and Matlab V7.1 SP3 Executable Specifications. Includes vugraphs, and Matlab code by Theresa Meuse of MIT Lincoln Laboratory. This release has been made to work with GNU Octave 2.9.9+ which is a Windows, non-Cygwin version of Octave (Octave 2.9.9+s URL is found in its README.txt file). PUBLIC RELEASE VERSION (toolbox-independent)
    • SSCA #4 - Simulation (TBD)
    • SSCA #5 - Simulation (TBD)
    • SSCA #6 - Simulation (TBD)

    SCCA Feedback/Comment Form

 

 © 2007 All Rights Reserved