The research work of this thesis mainly includes: Research of floating point algorithm, including addition, subtraction, multiplication, division, evolution and CORDIC ( COordinate Rotation Digital Computer). ·研究浮点算法,主要包括加减法、乘法、除法、开平方以及CORDIC(CoordinateRotationDigitalComputer)算法。
In the model data structure Pixel supports single-precision floating point, integer and other basic data types, away from general-purpose computer platform, the complex double-precision floating point data types, reduces the complexity of floating-point operations. 在模型数据结构中,支持单精度浮点数,整型数等基本数据类型,舍弃通用计算机平台中的双精度浮点数等复杂数据类型,减少浮点运算的复杂度。