Content deleted Content added
Line 163:
http://math.exeter.edu/rparris/peanut/cordic.pdf (page 16) gives a very weird way of computing square roots using trig functions -- anonymous
The exeter link seems dead.
The cordic cited method, I think (and will check), actually gives twice the square root and the error is copied from Table IV in the cited paper that is wrong. When Cordic is run in hyperbolic vectoring mode it computes sqrt(x^2 - y^2), so for x=s+1 and y=s-1 we get sqrt(s^2 + 2s + 1 - s^2 + 2s - 1) = sqrt(4s). [[User:David.J.Greaves|David.J.Greaves]] ([[User talk:David.J.Greaves|talk]]) 19:33, 22 August 2017 (UTC).
:We do have an article about [[CORDIC]], which mentions computing square roots but does not give further details. I added a sentence to this article with a link to CORDIC. I don't have an idea whether CORDIC is still used (or was ever used) to compute square roots, but it would be good to have some more details. -- [[User:Jitse Niesen|Jitse Niesen]] ([[User talk:Jitse Niesen|talk]]) 14:42, 8 December 2008 (UTC)
| |||