Why are we stuck with Zeros and Ones?

For one thing Bolean Logic would go into the dumper.

And to have 16 possiblities it may be more than that, such as 2 to the 16 power of possiblities? That would have to be more defined which would not be easy.

I like to give a practical example. Digital Electronic is a matter of gates.

I had a engineer that was working on a sorting conveyor that had (8) lanes, the customer wanted to have it programed to use 0, 1, or multiples of lanes simutanuously how he sold it and said no problem.

I pointed out to him that the possiblities were near impossible and showed him why, each gate affect the others. He did'nt pick up on it till he got to 64 possibilities and he did'nt even begin to touch the possibilities yet.

If there were to come up with a base 16 chip. that would be one powerful chip.

I'm not thinking 2^16....that's still base 2. Up it to base 16. Essentially redefining the "bit" to have 16 values instead of 2. More possibilities is my point. With a higher base, we can store and transfer more, ALOT more. Today, when we send 1KB of data, there are 8192 "slots" where each slot can be configured 2 ways (0 or 1). Imagine each bit can be represented 16 different ways instead of 2. Now there are 16 ways each of the 8192 slots can be configured.

2^16.... more 16^16 thats still not enough from your post on how you describe data transfer. I thought of that after I posted thinking heck I just nested base 2. but the possibilites could be huge, but defining the varied rates instead of just on and off. its more like going back to analog.

I think what you are thinking of is analogue processing. It is a lot more complex to design however we do have powerful computers to do that for us! There is the halfway method of colour coding, which is good for laser chips where different colours provide different channels of calculation each in Base 2. We have (not quite got) the technology.

Early fire control computers for battleships were analogue devices and mechanical as well!

In the 70's hp had PC and CPU's running on BCD (Binary Coded Decimal) The decimal digits were represented in nibbles (4 bit groups) in 64 bit registers. Calculation could be selected in the default BCD or binary.

The Assembly language for the 8086 in the XT for example contained the instruction AAA - (ASCII Adjust for Addition) which were essentially a BCD implementation. I think the function was implemented in Turbo Pascal by Borland.

The main disadvantage of binary representation was with fractions. any fraction was represented as the appropriate sum of 1/2 , 1/4 , 1/8 , 1/16 , 1/32 , 1/64 , 1/128. . . .

It was for example impossible to represent 1/3 accurately. And because of the limited number of bits the square of the square root of 2 did not return a result of 2.

Interesting,

processing power (speed) improve since the 70's.

Heck we're still solving for pi.

What about tri-state technology (-1, 0, +1) ?

That would also fit in with dual layer DVDs but it might not be such a great idea for magnetic storage on a hard drive..

Hendrik has hit the nail on the head.

In 1980 at university we were working with a hypotehetical 3 state logic (-1, 0 and +1) and developed source code and such for such a system.

We were doing theoretical bandwidth calculations and such and an 8 "bit" signal could have 3^8 possible states instead of 2^8 possible states.

As a theory experiment it was great, a little like the theory of transistors was way ahead of the physics/foundry capabilities to produce the necessary purity of silicon there was nothing physical that we could actually build.

Creating multi state software and such would have tremendous possibilities, but it obviously has not been persued (YET) but the quantum guys with positive and negative spin electrons might be getting there. (Or at least that was where we were heading with the thoughts at the time.)

Are you talking about fuzzy logic?

http://en.wikipedia.org/wiki/Fuzzy_logic

Tad

Ever considered basing on binary with the addition of the state between 1 and 0??

Without doing enough enough research to give a good answer, I'll just posit off the top of my head...Quantum computing...isn't that based on a base greater than 2?

You're not stuck with it at all. We are where we are today because of a paper written by Claude B. Shannon in 1937 considered to be the most important master's thesis in history. At Michigan, Shannon liked George Boole's work and later at MIT was encouraged to build a computer that could use binary algebra thereby replacing "analog". See the following link. He is considered one of the most important persons of the twentieth century. Claude_B. Shannon

I'm sticking my nose in where I know rather little. but isn't the whole purpose of a zero / one logic system so that one can change the "numbers" by simply turning the current on or off for that specific gate or chip component?

Well I am weak on the subject myself. but I believe the op is suggesting that instead of giving a 1 OR 0 output. to give an output from a multiple of possiblities instead of just 2. which would increase the processing power by making it more efficient.

I don't get it??

lol

Why did the chicken cross the road?

.sdeen s'daor eht ot evitnettani saw ti esuaceB

I suppose, at least, that's often how I manage to make my wife cross.

Hello coonj, Step analog is a underdeveloped computing system that shows much promise but has many hurdles to overcome. One is the digital thinkers.

To make a base 16 computer will take reinventing every aspect of the system. Buses get smaller giving more room on the chip but stacks and math get complicated.

All software and hardware will have to be scratch built. Emulators will be needed for binary software.

Clock speeds will start back in the KHz range and then move up as solution are found to new issues.

Big paradigm change for the people in the computer industry, don't expect acceptance very fast.

A few more years of speed limiting physics and it will be more viable way to process more information.

Brad

Hello coonj and other readers,

The fastest switching possible is 2-state: 1st state = No Power; 2nd state = power On.

Some 30 years ago, I did think about using the Trinary system, (+V, 0V, -V) as did others. With further thinking, the idea was dropped, as there are major difficulties for both code writers and hardware.

Likewise the use of "Voltage stepped" DC pulses to give hexadecimal or other "advantages".

The complexity for code-writers using that type of system beggars belief, and there appears to me extreme complications with hardware too, if that system is used.

It is sometimes not realised that the bandwidth requirements increase to a great degree, once the 2-state (0V, some V) system is not used.

The wonderful thing about the presently used Digital 2-state (0V, some V) system, is that once a threshold Voltage has been arrived at, the system considers that is a pulse = 1, regardless of distortion/s in waveshape, which often occur in high-speed systems.

Likewise the absence of a Voltage is the 0 pulse.

The gate and Software system is easily understood. I simply explain it to people who know nothing about electronics, by using the analogy of sheep-yards.

The sheep (Electrons) are herded in one end, and then the Farmer (Owner of the computer) gives instructions to the foreman (Operating System) and the yards are set up (The yards are temporary ones as per the coloured picture).

The Foreman then receives further special instructions from the Farmer (Computer Owner) because there are going to be several different things happen to the sheep (Electrons) and on the basis of those special Instructions (Use of Internet Browser, Spreadsheet, Document, any other Software) The Foreman places Men at the various gates, and each Gate Operator has a written Instruction Set from the "Special Instructions of the Software Program being used".

Next, the command is given, the sheep are moved through the yards, with various Gate Operators opening, closing, moving their individual gates.

Example: Farmer wants all sheep to go through the yards, and certain things to occur to achieve results.

The options chosen:

1. Old Rams to small enclosure for dog tucker
2. Young feisty rams to ram enclosure, later dipped and to the ram paddock
3. Pregnant ewes to separate enclosure, to shearing shed, then out for dipping, then off to ewe paddock near the house
   
4. Lambs aged 3 months or less, not to be shorn or dipped, then to go with their ewe mother to ewe paddock
5. Lambs aged over 3 months, to be placed in enclosure, then into truck, off to meat-works
6. There are many further optional actions, all selectable.

Sometimes a wolf may arrive, and disrupt the smooth flow, or another large creature attempt to destroy the sheep-yards as they have been set up for the special occasion.

In computer talk, this is a Trojan, virus, Worm or other malware, which interferes with the intended sheep (electron) processing.

Of course electronic pulses happen much faster than sheep move, but the principle is the same.

So, try and remember that the fastest switching state is 2-state (0v, some V) and any so-called trinary or other switching state becomes more complex, and uses thus more bandwidth (needs faster processor) as well as not being so easily stored.

2-state (0v, some V) logic is here to stay, folks, because it is the fastest and most reliable type of logic we know of.

Now tonight, remember not to lose count of those sheep there, as they successively jump the fence.

Kind Regards....

YAWN!

Enough with the sheep already, I'm getting sleepy

The semi conductor industry is working on a tristate transistor. There are lots of issues to finalize. A CPU which combines both base2 and base3 systems will be the results.