Point multiplication is a statistical acceleration scheme which first identifies a set of characteristic key points in a graphical calculation such as articulation or rendering. Samples of the calculation at the key points are then used to provide a subspace based estimate of the entire calculation. Point multiplication is a useful acceleration scheme when the calculation requirements for evaluating the key point values are substantially lower than for evaluating the full set of points. Calculations of this sort occur in many areas of graphics - here we present examples from the calculation of facial articulation and the rendering of scenes with global illumination.

The soft caching process is an extension to the point multiplication
technique where the key points are also used to provide
a confidence estimate for the point multiplication result.
In frames with high anticipated error the calculation will then ``fail through''
to a full evaluation of those points (a cache miss), while frames
with low error can use the accelerated statistical evaluation (a cache
hit).

*Additional materials:* [SoftCaching.mov]

Available as Pixar Technical Memo #06-04

Other versions: