一种80位扩展双精度浮点三角函数运算单元的设计  

Design of a 80 Bits Extended Double Precision Floating-Point Trigonometric Computing Unit

作　　者：张佳[1] 唐威[1] 盛廷义[1] 赵文琦[1] 艾刁 

机构地区：[1]西安微电子技术研究所,陕西西安710054

出　　处：《微电子学与计算机》2014年第4期23-26,共4页Microelectronics & Computer

摘　　要：三角函数求值这一运算计算过程复杂,硬件较难实现.针对这一问题,通过改进CORDIC算法,实现了兼容SPARC处理器和INTEL处理器浮点标准的80位高精度浮点三角函数的计算.在算法设计中,将函数的计算范围扩展至-π^+π,并且实现了迭代次数的可配置.最后验证了算法的正确性与完整性,分析了运算过程中迭代次数与精度的关系.结果表明,运算的精度提高到10-14.The trigonometric algorithm is hard to achieve. By improving the CORDIC algorithm achieved a 80- precision floating-point trigonometric calculation which compatibled with the INTEL processor and SPARC processor. In the design, the calculation range has been extended to -π ～ ＋ π, and the number of the iterations can be configured. Finally, the correctness and completeness of the algorithm are verified, and the relationship between the number of iterations and the accuracy was analyzed. The results showed that the precision was raised up to 10-14 , the simplicity and speed of the algorithm were also great advantages.

关 键 词：CORDIC算法 高精度 三角函数 浮点 

分 类 号：TN402[电子电信—微电子学与固体电子学]