论文部分内容阅读
介绍自主设计的龙腾C2微处理器中浮点运算单元的设计与实现。该处理器与Intel 80486DX4指令系统兼容,支持IEEE 754标准扩展精度的浮点基本函数和超越函数运算。介绍了浮点运算单元的结构,分析了实现超越函数的高精度CORDIC算法的流程,讨论了实现浮点超越函数运算的数据通路和控制通路结构,并给出了仿真结果和精度评估结果。仿真和分析的结果表明,浮点运算单元的设计满足龙腾C2微处理器的设计要求。
Design and Implementation of Floating-Point Computing Unit in Autonomous Longteng C2 Microprocessor. The processor is compatible with the Intel 80486DX4 instruction set and supports IEEE 754 standard extended-precision floating-point base functions and transcendental function operations. The structure of floating-point unit is introduced. The flow of high-precision CORDIC algorithm to realize transcendental function is analyzed. The structure of data path and control path to realize floating point transcendental function is discussed. The simulation result and accuracy evaluation result are also given. Simulation and analysis of the results show that the design of floating-point computing unit to meet the design requirements of C2C processor.