一、问题的提出 1.装配超快速小存的出发点×××计算机的运算速度是毫微秒(ns)的数量级,而其大容量主内存的存取周期是微秒(μs)的数量级,两者不相匹配。采用先行控制技术是解决此矛盾的一个办法。“指令堆站”、“先行取数寄存器”、“后行送数寄存器”等都是×××计算机实现先行控制技术的硬设备。但是由于系统的可靠性要求很高以及技术复杂、成本高等种种原因,这些硬设备的容量很小,只能容纳几个数(对于先行取数寄存器和后行送数寄存器)或几条指令(对于指令堆站)。因此不能使几十万至几百万大容量的主存跟上运算器的运算速度。所以在×××机先行控制技术中有必要装配存取周期为ns数量级的超快速小存作 First, the question raised 1. The starting point for the installation of ultra-fast small × × × computing speed is nanoseconds (ns) of the order of magnitude, and its large main memory access cycle is microseconds (μs) of the order of magnitude, The two do not match. The use of advanced control technology is a solution to this contradiction. “Command heap station”, “first take the number of registers”, “send the number of registers after” are all × × × computer control technology to achieve first-class hardware. However, due to the high reliability of the system and the technical complexity and high cost, the capacity of these hardware devices is very small and can only accommodate a few numbers (for the first take register and the next send register) or a few instructions For instruction heap station). It can not make hundreds of thousands to millions of large-capacity main memory to keep up with the computing speed of the calculator. Therefore, it is necessary to assemble an ultra-fast small deposit-taking cycle of ns order of magnitude in the advanced control technology of the ××× machine