There are always more optimizations that can be worked into code to squeeze more performance from the processor. Use of SIMD and XMM registers is the most important of them, but this chapter doesn't cover this subject which fills several chapters in other books (see Gerber 2002). This capability is important to explore anytime more than a single calculation needs to be performed. The benefits are substantial.

Other chapters in this book have also presented various techniques that deal with improving performance. Chapters 6 and 7, for example, discuss platform-specific performance sinks and solutions.