Content deleted Content added
added to project computing |
→Motive of the algorithm is unclear: new section |
||
Line 57:
[[User:Sudar 4edi|Sudar 4edi]] ([[User talk:Sudar 4edi|talk]]) 14:31, 18 March 2008 (UTC)sudar 4edi
:Yes, you can! [[Special:Contributions/124.30.235.62|124.30.235.62]] ([[User talk:124.30.235.62|talk]]) 07:12, 30 October 2008 (UTC)
== Motive of the algorithm is unclear ==
It is not clear to me why this algorithm is important. It does not seem to be an improvement over what one might do just by duplicating what I learned to do in multiplying decimal numbers by hand. Of course those are sign-and-magnitude numbers, but that can be dealt with by recognizing that in 2's complement numbers the high-order bit is a sign bit, and has the value of the most-negative number. Accordingly if that bit is on, it calls for subtraction rather than addition.
Following this idea, we get an algorithm that adds or subtracts once per 1-bit in the multiplier, shifts the same number of times as Booth's algorithm, and only has to examine one bit rather than pairs of bits. In software, this is perhaps easier (depending on the language). In hardware, I don't see much difference. Since this algorithm seems obscure, I'm wondering why it gets any attention.
Sorry to not cite references, but this is an extrapolation of what I learned in grade school. I'll add more advanced ones if it seems worth putting on the main page.
--[[User:Kogorman|kogorman]] ([[User talk:Kogorman|talk]]) 14:58, 18 September 2012 (UTC)
| |||