V.D.Baykov,V.B.Smolov, Hardware implementation of elementary
functions in computers, |
|
CONTENT
|
|
|
Convergence Analysis of CORDIC
|
The demonstration of
the “double iterations“ (radix=2)
necessity for arcsin and arccos functions
evaluation. |
”Double iterations” for ln and exp functions
evaluation. Fig.3
demonstrates, why need “double
iterations” for arcsin and arcos evaluation. |
|
|
Pay
attention: i=0,0, 1,1, 2,2,… mean : double
iterations!! |
”Double iterations” for the functions ln and exp
evaluation: i=1,1, 2,2, 3,3….. |
”Double
iterations” are necessary too for shX,
chX, ArthX |
|
|
|
Errors of CORDIC analysis |
|
Root-mean square error estimations of CORDIC |
Summary
Table of errors estimations for
Volder and Meggitt algorithms. |