It would seem impossible to make computers contain information much more densely than they already do. I have a solution to this problem. Rather than using just neutral and negative charges in computers, one could also use positive charges. 8 bits of binary could contain the same amount of information as about 5 bits of trinary, increasing computer information density by a factor of approximately 1.6.
-- apocalyps956, Jun 26 2006

US Patent 5,432,735: Ternary storage dynamic RAM http://patft.uspto....735&RS=PN/5,432,735
From 1993 - surprisingly recent. [jutta, Jun 26 2006]

You may want to do a search on quantum computing
-- theircompetitor, Jun 26 2006


I've thought about this a few times in the couple of years that I've known about binary.
-- BJS, Jun 26 2006


in howard taylor's schlock mercenary trinary code is mentioned in an annotation. the three possible states he mentions are yes, no, and maybe.

i admit he makes no claims to his authority in terms of computer programing languages, as well as his other sciencey claims.
-- tcarson, Jun 26 2006


Maybe the "maybe" could be random?
-- BJS, Jun 26 2006


We have pondered this at great length before.
-- DrCurry, Jun 26 2006


<pedant> ternary? </pedant>
-- pertinax, Jun 26 2006


So, would we be toggling not bits but --
-- jutta, Jun 26 2006


[jutta] taylor also mentioned how a whole generation of computer programers didn't procreate because of their use of tit in talking about trinary code.
-- tcarson, Jun 26 2006


How do you titillate an ocelot? -- You oscillate its tits a lot
-- xaviergisz, Jun 26 2006


Why stop at three? If this is useful you could use a number of voltage states (four, five, six ...) to reduce the number of bits further. It's just a matter of how you interpret the voltage levels.
-- NoOneYouKnow, Jun 26 2006


//Why stop at three// My computer science teaching guy proved to us that three states has a higher efficiency of data storage than four or five. In fact, e is the most efficient, but 2.7182 states is impractical.
-- phundug, Jun 26 2006


Doesn't work. There are no "Positive Charges", just lack of negative ones.

Any bit of data storage is basically a bucket for electrons. It either has electrons in it, or it doesn't. On and off. 1 and 0. Binary.

Anything else is a complete, from scratch reinvention of computing. See "Quantum"
-- Galbinus_Caeli, Jun 26 2006


Dont confuse date storage algorithms and Computer processing language. Galbinus is dead on as far as the eficacy of this system. all the aforementioned points are useful only for storage of information(a .5 volt charge followed by a 1 V charge is the equivilent of 0110101010010101 01010001010101) thats doesnt mean you dont need the binary you just dont need to store it verbatim, still need to decode and run the binary for it to work)

This is bad science. (-)
-- jhomrighaus, Jun 26 2006


In theory this sounds like a reasonable idea (if you assume three or more voltage levels, rather than a positive, neutral , negative)...

but in practice...I would guess that the storing, retrieving and the corresponding logic gates and buses would be more complex to implement and would take up more space.

Also - the more levels you have, the more chance you have of a value being 'mistaken' for another level

[-] sorry
-- monojohnny, Jun 26 2006



random, halfbakery