对数函数的FPGA怎么实现

嗨，大家好...
我正在寻找对数函数的FPGA实现。
大多数论文建议基于LUT或LUT和插值相结合。
这些方法最多可提供20位的精度。
实际上我要求双精度，所以这个精度很差。
然后我通过xiliix的IP记录获得了douoble精度，它提供了近53位的精度。
在这里，我没有得到xilinx正在使用哪种算法或哪种方法，这提供了这么多的准确性
是基于staylor系列..？
如果是，那么那也将有更大指数的术语不是它（如x ^ 2，x ^ 3 .... x ^ 20 ......）
任何建议都有帮助
谢谢



以下为原文

HI everyone...

   I am looking for FPGA implementation of logarithmic function. Most of papers suggesting either LUT based or LUT and interpolation combined one. These methods will give accuracy maximum upto 20 bits.
Actually I am going for double precision so this accuracy is very poor.
  Then I gone through IP of Log from xiliix for douoble precision which gives accuracy of almost 53 bits. Here I am not getting which algorithm or which method xilinx is using which gives this much of accuracy

is it staylor series based ..? if yes then that also will have terms with larger exponent isnt it (like x^2, x^3....x^20 ......)

any suggestion are helpful

thanks