An Improved Vertex Caching Scheme for 3D Mesh Rendering

Gang Lin, Thomas P.-Y. Yu

ECSE & Math Department

Rensselaer Polytechnic Institute


The K-Cache-Reorder (download here) program generates a rendering sequence for the input mesh using greedy optimization to minimize vertex cache misses during rendering. The output model is the same as the input model except that its faces are reordered. A few examples of our test models: Venus, Bunny, Horse.

Arguments:          (1) input file: a polygonal model in VRML text format
                             (2) cache model: 1 (for FIFO) or 2 (for Controllable)
                             (3) cache size
                            (4) output file: the model in VRML with reordered faces

Usage: K-Cache-Reorder <input file> <cache model> <cache size> <output file>

Example: K-Cache-Reorder c:\temp\bunny.wrl 1 12 c:\temp\reordered_bunny.wrl

For detail of the algorithm, please refer to paper (PDF)

Please send mail to ling@ecse.rpi.edu for any questions or comments. Thank you!