|
Written by Matthias Mann
|
|
This is a java code snippet that can be used to optimize geometry for GPUs.
It's based on the paper Linear-Speed Vertex Cache Optimisation from Tom Forsynth.
My implementation is just a straight forward implementation which has a O(nē) runtime performance. Futher optimisations are left for the experienced reader.
The source code can be downloaded as reference (MIT license) - it is an extract from the worldscape engine. I removed all dependencies on other worldscape code. I'm not even sure if it compiles right out of the archive - but it should give you a base to build something useful with it.
If you found a way to improve the performance I would be thankful if you publish your enhancements and drop me an email. |
|
Last Updated ( Wednesday, 30 May 2007 )
|
