The Algorithm Design Manual
About the Book
Programming Challenges

The Stony Brook Algorithm Repository

Steven Skiena
Stony Brook University
Dept. of Computer Science

OpenCCL: Cache-Coherent Layouts of Meshes and Graphs


OpenCCL is an open source code to compute cache-oblivious layouts of graphs and polygonal meshes to improve performances of a wide variety of applications including geometric processing and visualization applications. OpenCCL has following features:

* Generality: OpenCCL can be used to compute cache-oblivious layouts of any graphs and any polygonal meshes including hierarchies and multi-resolution meshes. * Cache-Obliviousness: OpenCCL requires no knowledge of cache parameters or block sizes of the memory hierarchy involved in the computation. * Performance: Applications can experience significant improvements on performances by using cache-oblivious layouts of OpenCCL. For example, we observed 2 - 20 speedup on performances on isocontour extraction, view-dependent rendering, and collision detection. However, there is no guarantee to improve performances of any applications. * Usability: Cache-oblivious layouts computed by OpenCCL can improve performances of applications without modifying any algorithms and runtime applications.


  • Download Files (local site)
  • Offical Site

    Problem Links

      
    Graph Data Structures (7)



    This page last modified on 2008-07-10 .
    www.algorist.com