论文部分内容阅读
乘法是日常生活中经常遇到的一种算术运算。用计算机实现乘法运算的最简单的方法是加和移位算法,除了数据格式是用二进制表示以外,它和普通的笔算非常类似。图1示出了一个8×8乘法的例子。每一位乘数,都产生一个部分积,然后相加。对计算机来说每得一部分积就要进行一次加法和一次移位操作。对n×n位乘法,则需n次加和移位操作。实际上,在通用计算机上用软件实现乘法还要复杂些,速度甚慢。以Intel 8080微处理器为例,调用一次乘法子程序,完成一个8×8乘法运算,约需250～300μs时间。对于许多科学计算和实时数字信号处理,这是远远不能满足要求的。例如在矩阵运算、数字滤波和FFT(快速付
Multiplication is an arithmetic operation often encountered in everyday life. The easiest way to do multiplication with a computer is to add and shift the algorithm, which is very similar to an ordinary one except that the data format is represented in binary. Figure 1 shows an example of an 8 × 8 multiplication. Each multiplier produces a partial product and then adds. For computers, each part of the product will be an addition and a shift operation. For n × n-bit multiplication, n additions and shifts are required. In fact, multiplication in software on a general-purpose computer is more complex and slow. To Intel 8080 microprocessor, for example, call a multiplication subroutine to complete an 8 × 8 multiplication operation takes about 250 ~ 300μs time. For many scientific computing and real-time digital signal processing, which is far from meeting the requirements. For example, in matrix operations, digital filtering and FFT (fast pay