Fast GPU ray tracing of dynamic meshes using geometry images

Using the GPU to accelerate ray tracing may seem like a natural decision due to the highly parallel nature of the problem. However, past attempts at a fully general GPU ray tracer have yielded poor performance compared to optimized CPU ray tracers. In this paper we instead focus on an algorithm for quick intersection of a single mesh on the GPU. The algorithm uses a threaded bounding volume hierarchy built from a geometry image, which can be efficiently traversed and constructed entirely on the GPU. For single meshes this acceleration structure beats previous methods by an order of magnitude, and allows for both dynamic geometry and an efficient level of detail scheme at no extra cost.