Math chips: How they work: Augmenting microprocessors, they speed up math operations while giving systems designers a variety of performance, cost, and integration options

The use of mathematics coprocessors to speed up calculations on desktop computers and workstations is discussed. The reasons why interest in speed of mathematical operations rather than in the rate of program execution has arisen are examined. The different types of coprocessors are described. The different types of coprocessors are described, and the construction from standard building blocks of numeric processors that execute code independently of the main processor, thus offering even faster performance, is discussed. General-purpose attached numerics processors, which contains elements of both approaches, are described. The tradeoff between ease of use and performance is considered.