Square Root Algo

I googled "determining square roots by recurrence fpga" and found some articles.

http://www.springerlink.com/content/q1l7525484226550/

http://ww1.ucmss.com/books/LFS/CSREA2006/CDE4842.pdf

I didn't go through all the hits - you should. But, don't believe everything you read on the net, or even in books. I spent a couple of weeks researching cyclic redundancy codes recently and was amazed by all the misinformation out there.

Good luck with this, it sounds like fun.

Thnx for all your effort friend! I will definitely check those links but I assure you that I have spent almost a week looking for some material, but all I found was an IEEE paper by M. D. Ercegovac and T. Lang fathers of this algo and nothing more. One more thing I am working on the Digit Recurrence Algo, any other algo won't do. Thnx again.

"Digit Recurrence Algorithm" and "Square Root Algorithm" used as a search term in both <Google.com> and <Dogpile.com> will keep you busy all night.

When doing square roots, I have used Newton's method. You can use tricks for the divide so you don't have to do that. You can get great bit resolution when using this method.

"When doing square roots, I have used Newton's method. You can use tricks for the divide so you don't have to do that. You can get great bit resolution when using this method."

Yes I have read Newton-Raphson method and it is the most widely known tech to evaluate roots. I switched over to Digit Recurrence because of its high efficiency in terms of area reqd for implimentation. The Newton method requires large chip area for the look-up tables as compared to digit recurrence. Also evaluation of floating pt. nos. is more efficient with the digit recurrence.

Look up CORDIC.