Computers are everywhere, and they all communicate through a system of numbers called "binary".

Binary's basic unit is a bit, which has two states of being: on or off.
On=1
Off=0

A transistor is a small piece of machinery like a switch that is switched electronically to either on or off, and the computer reads this as a bit. "Flash memory", the memory used on Flash Drives, MP3 Players like iPods, and memory cards, is simply a bunch of transitors on a chip that the computer controls to put information on.

After the introduction of the transistor, a man named Gordon Moore predicted that the number of transistors that could fit on a chip would double every 12-16 months due to improvements in manufacturing techniques and other factors.

This turned out to be true, and so every year or so flash drives get more space for the same price, and more memory is available as RAM in computers. This trend became known as "Moore's Law".

(As a side-note, the graphics cards produced by nvidia occasionally break this law, by doubling in only 10 or so months)

Transistors, having two states of being, are perfect for use as bit representation. Eventually, however, due to the uncertainty principle, we will reach the limit to how small we can make transistors. We can't go to atomic scale, so we'll reach the end of the rope eventually.

What is the solution to this problem?

It's simple.

The nature of the binary system implies that only two values, on and off, are possible within the system.

By adding another value to binary, (and thus renaming the system "trinary"), we could solve the problem.

The issue here is that we can't have trinary without a piece of machinery that has three possible states instead of the two that transistors have.

This means making transistors with a third state, perhaps named "Neutral", and as an effect allowing for even more information to be stored on the same size chip. A 2 gigabyte chip would be able to hold 3 gigs, a 4 gigabyte could hold 6, etc.

Discuss.

Sounds like a good idea in theory. I have no idea how you would go about doing it.


Also I don't see what there is to debate?

Quote from MadHatter

Also I don't see what there is to debate?

ditto, looks like like a discussion.

But I don't get how that works. On simple, large-scale terms: When you, say, press a button, there are "Yes it is pressed" and "No it is not"; adding a "maybe it was pressed" wouldn't solve much.

I think I don't know how computers work. Explain.

Quote from Bonk

But I don't get how that works. On simple, large-scale terms: When you, say, press a button, there are "Yes it is pressed" and "No it is not"; adding a "maybe it was pressed" wouldn't solve much.

I think I don't know how computers work. Explain.

Instead of a button with two positions, think of it as a switch with three positions.

Nice idea but then wouldn't all binarly would have to be re-writen to include this new value since it would be in the middle with the originals as extreams of each side? Its a clever idea but if we get to the stage of being unable to make transistors smaller wouldn't we just make the chip bigger, screens will (probably) always be quite large and you could fit alot in by the time their as small as possible.

Quote from Scorpioxxx

Nice idea but then wouldn't all binarly would have to be re-writen to include this new value since it would be in the middle with the originals as extreams of each side?

Kind of.

This would require completely new methods of wiring and programming, but in the end we could fit much more code and images into far smaller computer memory.

Its a clever idea but if we get to the stage of being unable to make transistors smaller wouldn't we just make the chip bigger, screens will (probably) always be quite large and you could fit alot in by the time their as small as possible.

No, that wouldn't be progress. We already space to as much effect as we can, so by the time we get to the end of the rope in our transistor cramming abilities, our chips will already be just as big as the screen.

To explain, I'll use the Playstation Portable as an example. if you were to open it up, you'd see that the entire profile of the console's inside is used by the PSP's circuitry. If we kept the PSP at the same size for the PSP 2, 3, 4, and 5, but increased the number of transistors on it's chip, then eventually we would end up reaching the limit that we can use the space. If we improved the PSP by making the PSP 6 bigger than the PSP 5 to accomadate mroe transistors, that WOULD be an improved PSP, but it would be bigger. This would be fine for 1, 2 or even 3 versions, but the console would continue getting so big and heavy that sales would droop no matter how great the console was.

Making the chips bigger isn't a solution, it's a workaround.

I don't know much about transistors, but wouldn't a third switch added to a two switch system just be a middle ground between the original two? That doesn't really seem like potential for more memory storage.

Yeah indeed. I was taught that transistors need to be either ON or OFF, a "I DONT FUCKING KNOW" option seems rather pointless.
Whereas binary with the old-fashioned two positions is working perfectly fine and would therefore not need to be changed in any time soon. Changing the system every computer basically depends on would end up in a huge mess.

Do I REALLY need do explain this mathematically so you can understand?


Currently, ASCII characters are stroed on a computer in binary units of 8 characters called a "byte".

The character "1" is stored as "00110001", and "a" is stored as "01100001"

B is 01100010
C is 01100011

The list goes on.

In binary, each 1 and each 0 takes up an entire transistor, and every byte can only be one of 256 combinations, with a byte being 8 digits.

That means that to store the word "Binary" (011000100110100101101110011000010111001001111001) you need 48 transistors.

With a trinary system, you have a whole other symbol to use (0, 1, and 2 rather than just 0 and 1) and so in an 8 character byte with 3 possible characters instead of just 2, you can have a whole lot more combinations.

Because we don't even need that many combinations, we can store ASCII characters with only 4 digits to a byte, and therefore only 4 transistors.

If we store the word "binary" in 4 character bits with trinary instead of binary, it would look something like this:

010220211020102221120111

Count the digits. That's only 24 digits. If each digit is stored on a single transitor with "off", "on", and "neutral" positions, then we only need 24 transistors to store the exact same amount of information.

Imagine a text document like my post beiing stored in binary. It would need 13288 transistors to be stored in ASCII. However, if stored on trinary, it would only need something like 4429 three-state transitors.

So, to close, code can be stored in a trinary system in a whole lot less memory space.

The real issue is figuring out how to make the trinary transistor.

I think another consideration would be whether or not we actually need to store more information by the time we reach this limit. I have over half a terabyte of memory on my computer in a pretty small amount of space.

From a realistic viewpoint, there's no need to store hundreds of terabytes on personal computers. Hell, even one or two is excessive for most people's uses. And so far, we've been able to keep up with large-scale memory requirements.

edit: alright, that makes more sense Ash. And yeah, explaining it mathematically would be nice if you're discussing a topic not everyone knows a lot about.

My porn folder is 3 terabytes, what the **** are you talking about. We need more.

I understand what you're saying, but having a third option for a yes or no question or a light switch doesn't make sense.

yeah, cause i'd rather like to know that the light MIGHT be on/off, rather then knowing its on/off.

having that would be ****ing useless.

That's what I'm saying. It doesn't make sense. True or False is all you need.