Any electronic storage devices using more than 2 variables?

[The Grimmus]

I was day dreaming in my math class agian and somehow got thinking about binary and if instead of Trues's and falses or 0's and 1's
it where to be soemthign like 0's 1's 2's 3's and 4's like an example

0 would be = to the binary pair 00
and 1= 01
3=10
4=11
so writting the letter a would go from

01100001 to 1301
thus increasing the storage space 2 fold (i always wanted to use that but never could)

 

[enigma]

They don't have them yet, but that is the gist behind quantum computers which may be coming on the market in the next few decades.

 

[Lyuokdea]

What you are referring to is analog technology, anything that is analog, such as the analog joystick or your video tapes run under such a system, there are two reasons though, why a binary system works better for electronic computers.

1.) Stability - It is much easier for the computer to tell between two opposites then four or 8 (most analog devices use 8) different types of data. For instance on data CD's, when they are burned, there are very small microscopic pits, which look as so

[ ] [ ] [ ]

those pits can either be burned, so that they don't reflect light or not burned so that they do, (which one reflects light i don't remember for sure, it might be the other way around) it is much easier for a computer to tell if it is or isn't reflecting light then trying to figure out how much light is being reflected when it makes its read, that is why you can copy Mp3's millions of times across the internet and the sound is just as good, where if you make a copy of a copy of a copy of a video, it will look like a copy of a copy of a copy of a video.

2.) Speed - This might not make sense at first, because you would think allowing four different sets of data would increase speed, but because of the laws of physics the opposite is actually true. There is a set time which you must wait in between sending electronic signals because of a physical law, (I can't remember the name of it) that governs the time which is required for the voltage in the wire from the last signal to dissipate. When you run four or eight distinct voltages, which must be spaced apart enough that they can all be interpreted from each other, the time required to go from the highest voltage to the base voltage is very great, and thus signals cannot be sent very often through the wire. However with binary data, the two opposite voltages can be places very close, and the base voltage can already stand one of the numbers, in binary data it stands for zero, thus, signals can be sent at a much faster rate.

Although your idea is not practical for binary computers, it is however practical, and will be used for newer systems yet to develop, both quantum and laser driven computers will use analog systems as they do not have to obey the laws of the voltage transmission in a wire.

P.S. if anybody knows the law for the wire voltage and its equation off of the top of their head, I would like to know what it is. Thanks

 

[Heathcliff]

Computer buses use tri-state logic. High, low, and high impedance.

http://ourworld.compuserve.com/homepages/g_knott/elect357.htm

 

[Guybrush Threepwood]

Originally posted by Heathcliff
Computer buses use tri-state logic. High, low, and high impedance.

yes but that's because you can have multiple sources on the bus.
For storage it's different. You can't read high impedance, you can read only 1 or 0.

 

[wimms]

Analog computers were a blast. They were damn fast and damn precice. the only problem was with programming them - it came down to rebuilding your "pc" from "scratch"..

Every xDSL modem is encoding multiple bits into clock cycle, its called something like quadrature modulation. Afaik they have gone as far as 64 values per cycle, maybe even more.

Afaik Gigabit ethernet over copper uses many bits per clock.
I'm not sure if they use this, but any magnetic media can benefit from this, thus hard disk drives.
Even DDRAM chips could be built, as bits are stored in charged capacitors, but I believe there are lots of problems associated that proved this approach not forth it.

 

[The Grimmus]

Originally posted by enigma
They don't have them yet, but that is the gist behind quantum computers which may be coming on the market in the next few decades.

What will be the diffrence between them and regular binary storanges?

 

[Heathcliff]

"Analog computers were a blast..."

http://dcoward.best.vwh.net/analog

 

[Dissident Dan]

Originally posted by Lyuokdea
What you are referring to is analog technology, anything that is analog, such as the analog joystick or your video tapes run under such a system, there are two reasons though, why a binary system works better for electronic computers.

1.) Stability - It is much easier for the computer to tell between two opposites then four or 8 (most analog devices use 8) different types of data. For instance on data CD's, when they are burned, there are very small microscopic pits, which look as so

An analog device, by definition, couldn't use 8 values. That would still be digital, but octal (or octary?) digital instead of binary digital. When you have discrete, distinct values, that is digital. The defining element of analog is that values are continuous.

The primary problem with analog that I'm aware of is that you always have a degradation of signal, which means that exact copies are impossible. With digital, as long as the signal doesn't degrade by a huge amount, you can get the right value.

You experience this effect when you make copies of cassette tapes. The store-bought cassette is always better quality than a copy you made of it. This degradation just builds up when you make a copy of a copy of a copy. For example, if you retain the original signal to 98% accuracy, the signal accuracy after 15 copies is 0.96¹⁰ = 54.2%. That is unnacceptable for many applications.
-----------------------------------
Now, as far why we don't have many values for digital:

It would require some combinations of a wider spectrum of voltages and more precise measurement, which would use more energy and/or be more expensive.
Also, our computers are based on AND, OR, and NOT gates, which operate on the principles of true and false. I don't know if you can expand the concept to accommodate multiple values. For example, what would be the difference between (3 AND 2) and (1 AND 4)? In any case, the circuitry to deal with it would probably be more more complicated, negating the transmission speed benefit.
--------------------------------------
What is it about analog that allows parallel processing?