The problem of computer addition comprises two distinct tasks: (1) the design of an integer adder, and (2) its application to the addition of floating-point numbers. Integer addition is a nontrivial but relatively simple operation that can be performed for ordinary bit-widths within a single clock cycle. Consequently, this operation is generally treated by the RTL designer as a primitive operation to be implemented by a logic synthesis tool. This amounts to a selection from a library of pre-defined adder modules, based on width and timing requirements. Thus, our treatment of this topic, Sect. 8.1, is limited to a brief introduction, focusing on basic concepts that are relevant to later chapters. The remaining two sections of this chapter deal with optimization techniques that are commonly used in the normalization and rounding of floating-point sums.