A workload made of fine structures
Hair is not simply a dense triangle mesh. Large numbers of fine strands create difficult visibility and execution patterns. Grouping those strands and choosing how the GPU processes them can substantially change the shape of the workload.
The images on this page were rendered on a mobile GPU, which is the part of the claim that matters. Strand rendering at this quality is not remarkable on a desktop part with the power and bandwidth to spend; doing it inside a mobile budget is a different problem, and it is the one this work was about.
Clustering and execution
That work included capsule clustering and comparisons of global-pixel and groupshared execution paths. We measured how the organisation and size of the workload affected GPU execution time.
This is implementation-level research: a useful answer needs the data representation, GPU path and measurement conditions together. It also needs to preserve the visual behaviour that the application requires.
