What’s New

With the recent adoption of new APIs such as DirectX® 12 and Vulkan™, we are seeing renewed interest in an older tool.  AMD Tootle was originally developed in 2006 based on research published at I3D & SIGGRAPH.  As part of AMD’s GPUOpen initiative, AMD Tootle is now open source.  With this release, we have updated AMD Tootle to build with Microsoft® Visual Studio® 2012-2015 on Windows® 10.  On Linux®, it was tested on Ubuntu 15.04.

In case you are wondering, the name Tootle is inspired by the pronunciation of T. O. O. Tool (Triangle Order Optimization Tool).

Why Tootle?

AMD Tootle can be easily integrated as part of a rendering or mesh pre-processing tool chain since it does not require any update to the rendering logic. Yet, there is a potential high reward through significantly improving the rendering speed of graphics applications that are limited by vertex or pixel performance.  This improvement is gained by reordering triangles (both vertex and index buffer) prior to rendering to optimize for the post-transform vertex cache, early Z-culling hardware and hardware cache line size in the GPU.  This mesh optimization is applicable indiscriminately for any modern GPUs.  After optimization, the resulting triangle mesh can be saved out to disk to replace the original model for rendering as necessary.

To learn the details of AMD Tootle algorithms, I encourage you to check out the two AMD Tootle papers and talks included in the Tootle’s GitHub repository.  This repository includes Tootle library source code to be integrated into a mesh processing system and a sample Tootle command line tool utilizing the library.

Because there is little drawback to using AMD Tootle with a high potential reward, there is no reason not to optimize your 3D models.  So have you tootled your models?