处理器中的浮点除法和平方根运算算法

Floating-Point Division and Square-Root Arithmetics on Processors

硬件设计中发展了许多除法运算算法,各算法在商收敛性速度、基本硬件单元和数学公式等许多方面均不相同。通过对现在较流行的浮点除法和平方根运算算法进行介绍,分析各浮点除法和平方根运算算法的思路和适合的不同场合,比较各自的优缺点。举例说明LSFT32处理器中浮点除法算法的选择。只有当算法的思路及其特点与运算器的结构相匹配时才能充分发挥速度和规模的优势,所选用的算法才是有意义的。

Many algorithms have been developed for implementing division in hardware. These algorithms differ in many aspects, including quotient convergence rate, fundamental hardware primitives, and mathematical formulations. The paper presents a study of floatingpoint division and square - root arithmetics. It analyzes the main idea and environment of each algorithm, also compares the advantages and disadvantages between several prevalent algorithms. This paper takes the LSFT32 processor for example and shows how to choose a suitable algorithm. While the ideas and characteristics of the arithmetic are matching to the arithmetic logical unit, high speed and small scale will be developed. The choice is just meaningful.