This book describes numerous methods for using (and misusing) existing features in graphics hardware to implement and accelerate advanced shading algorithms. But graphics hardware is a fast-moving area and the features that were available at the time of this writing may be outdated by features available to the reader. This chapter provides some context on both what our assumptions were about the direction shading hardware was going, and how the concepts of the book will apply and adapt to this brave new world: the ever changing present.

Over the same time frame covered in this chapter, general-purpose CPUs will also likely evolve. It may even be that graphics accelerators and CPUs will become indistinguishable. However, at the moment, GPUs have certain specialized features, such as a deep texture prefetch pipeline and hardware implementation of rasterization algorithms. These specialized computational resources give them a large performance advantage over general-purpose CPUs for graphics operations that will probably persist for at least the next few years.