So, here's my dilemma:
Normal boolean logic has two sates: true and false.
In a computer they are represented by 1 and 0.
I'm working on a computer that will, instead of 1's and 0's, use
--'s, +'s, and |'s
Would -- be false, | be true and + be both? Does that even make sense?
Or should I leave boolean logic behind and make something new?
I've nearly gotten the basics of this computer figured out. (The only thing I've got left is a pesky little problem with the 'and' gates.) However, I haven't gotten the logic it will run on figured out yet.
This has stumped me... Help please!
No, this is not a quantum computer (at least not in the traditional sense). And no, I won't give many details about the actual mechanics of it's operation.
| ocalhoun wrote: |
So, here's my dilemma:
Normal boolean logic has two sates: true and false.
In a computer they are represented by 1 and 0.
I'm working on a computer that will, instead of 1's and 0's, use
--'s, +'s, and |'s
Would -- be false, | be true and + be both? Does that even make sense?
Or should I leave boolean logic behind and make something new?
I've nearly gotten the basics of this computer figured out. (The only thing I've got left is a pesky little problem with the 'and' gates.) However, I haven't gotten the logic it will run on figured out yet.
This has stumped me... Help please!
No, this is not a quantum computer (at least not in the traditional sense). And no, I won't give many details about the actual mechanics of it's operation. |
http://en.wikipedia.org/wiki/Ternary_logic
That might help. Usually when you have a 3-state logic, it's referred to as true, false and undefined, or +, - and 0. | Indi wrote: |
| ocalhoun wrote: | So, here's my dilemma:
Normal boolean logic has two sates: true and false.
In a computer they are represented by 1 and 0.
I'm working on a computer that will, instead of 1's and 0's, use
--'s, +'s, and |'s
Would -- be false, | be true and + be both? Does that even make sense?
Or should I leave boolean logic behind and make something new?
I've nearly gotten the basics of this computer figured out. (The only thing I've got left is a pesky little problem with the 'and' gates.) However, I haven't gotten the logic it will run on figured out yet.
This has stumped me... Help please!
No, this is not a quantum computer (at least not in the traditional sense). And no, I won't give many details about the actual mechanics of it's operation. |
http://en.wikipedia.org/wiki/Ternary_logic
That might help. Usually when you have a 3-state logic, it's referred to as true, false and undefined, or +, - and 0. |
I don't know if that is what he is looking for but from this angle you seem to be right on. Interesting concept but I think that the "gate" issue will prove to be more of a formidable obstacle than likely given credit for. On and off is easy, but when you get in to some other state than true/false(on/off) you get into a more complex beast as a whole. To represent this logic there has to be some sort of signal to transfer the information.
The complexity from this mere third bit of information isn't so easily brushed off as a detail left to be polished off. This is not to say that the logic isn't possible or even highly feasible. I just cannot help but wonder if the pay off, at a hardware level, will be worth the trouble when the work performed is matched against the traditional boolean logic of machine language. Although the machine logic is composed of two, the logic made from these components are much deeper, complex and more multi-faceted and flexible.
What is the motivation (payoff) for the use of this 'ternary' logic? Re-invention of the wheel or mere experimentation with complex systems? | eznet wrote: |
What is the motivation (payoff) for the use of this 'ternary' logic? Re-invention of the wheel or mere experimentation with complex systems? |
Theoretically, the payoff would be that one ternary byte (if I stick with an 8 bit byte) could hold the information of 25 binary bytes.
One binary byte: 256 possibilities
One ternary byte: 6561 possibilities
Also, a CPU using this system would have a much wider range of abilities, probably including being able to do more complex operations in a single cycle. Also, the medium that this data will be traveling through will produce far less waste heat, so it has a great potential for smaller sizes and faster clock speeds.
Also, thanks for the link, that is very interesting, and a little bit helpful. | ocalhoun wrote: |
So, here's my dilemma:
Normal boolean logic has two sates: true and false.
In a computer they are represented by 1 and 0.
I'm working on a computer that will, instead of 1's and 0's, use
--'s, +'s, and |'s
|
Traditionally, tri-state logic is regarded as a traditional binary logic gate with an extra 'enable' input. The symbol is :
or
Regards
ChrisI don't know it by heart but we learnt on university that most efficient algebra for computer would be one with base 'e' that is 2.7xx they choose base 2 because it is easier to implement but 3 is closer to 'e' that means that it would be more efficient
^How the heck can you have a logic system based on 2.7xx values?
What is .7 of a value? (false=1 of a value fal=.7 of a value?)
Not 2.7xx values, e values. As in the transcendental number 2.71828 18284 59045 23536 (to 20 decimal places).
Though I can't see what makes that more palatable to a computer, thats computer science more than I care to learn about.
Personally a system with YES, NO and MAYBE would make a more amusing logic system...
i think 3state logic is meaningless cause of the upcoming quantum pc's.. u'll waste a lot of time for nothing..
^Quite the opposite. If it works properly, it'll make quantum PC's obsolete. (Though it will work much more like a normal PC than a quantum PC would.)
It will also be much easier to manufacture, and be made of much more common materials.
One of my next experiments will be to see how much higher the clock speed could be (above a normal electronic computer).
No sarcasm intended, I'd like to see it work.
If you (I mean the general you, "one") could define a logic system based on e, I think it would be too difficult for humans to comprehend or program, so it would be useless.
But I'd be interested to see if your 3-state computer can be physically implemented and whether the logic is easy enough for humans to grasp.
^If all else fails, I could use it as a two-state system, and ignore the greater possibilities offered by three states. It would still be much faster, I think.
You could argue that three-valued logic is alreay in common use. In SQL databases, you can distinguish between true, false and null. In PHP there's $x==true, $x==false, and not isset($x);
