论文部分内容阅读
迭代除法已在高速流水线运算器中广为采用。但由于它在求商时不留余数,所以,一般认为迭代除法对求双精度商是很不方便的,不得不保留‘保留余数’除法指令。本文针对这一问题作些初步探讨,认为迭代除法不仅具有计算速度快之特点,还具有快捷求双倍位商的优点,而且费用低廉,只需在原有迭代除法基础上稍加改进,就可以得到一种软、硬件结合的双精度除法。
Iterative division has been widely used in high-speed pipeline calculators. However, since it does not leave a residue when it comes to finding a business, it is generally considered that iterative division is inconvenient for finding a double precision operator and has to retain the division instructions of “Remaining Remainder”. In this paper, some preliminary discussions on this issue, that the iterative division not only has the characteristics of fast computing speed, but also has the advantages of fast and double bit quotient, and low cost, only the original iterative division based on a slight improvement, you can Get a combination of software and hardware double precision division.