I'm gonna get flamed to no end for this but...

[oolceeoo]

Hey everyone. I don't know if anyone remembers my question a few months ago I posted on here about a theory that I have. It basically was that I believe that computers can exist without memory.

I want some input as to what I should do with my testing of this theory. I fully expect that if I do get any responses, I will be called stupid, go home kid, this is not a place for homework, idiotic, etc. But I hope that maybe atleast one reader will take me seriously.

I am putting my theory to the test currently that computers can indeed function without memory. What I am doing is getting very old computers and purposely overheating them with a heat gun. Why am I doing this? Well, I've seen video controllers overheat and produce artifacts on the screen. These artifacts are what you would consider a computer 'malfunction'. I propose that a computer malfunction, aside from software bugs, could possibly be the computer operating 'on its own'. A malfuction would be something that is not doing what it was intended to do.

Since computer circuitry that gets too hot is permanently damaged and will die prematurely, the key is finding the right balance of heat that can produce these artifacts without damaging the circuitry. This is proving to be very difficult, and since I don't know of anyone doing something like this, I don't know how to do it yet either.

I know that this all sounds crazy and a big waste of perfectly good computers, but I feel as though this could lead to something big. Please post some constructive criticism, and not just 'you are an idiot.' i get that enough each day.

 

[endyen]

I see what you are saying. Any random result could be considered as a result of a higher order.
I read a story about a guy who randomized his life, by the use of dice.
A little too out there for me though.

 

[MacCleod]

A human-example of what you're talking about. Consider what kind of 'computing' you would be able to do if you couldn't remember more than 1 second of your life at a time. Perhaps you could do simple math, but you'd have no idea that the reason you were doing it was to file your tax return, finish a test, etc. You couldn't do it.

Take it one step further. You're computing the sum:
43 + 89. First of all, where do you store the numbers 43 and 89? Where do you store the fact that you want to add them? When you add 9 + 3 to get 11, where do you store the 1 in the ones place, and the 1 in the ten's place?

Basically, I'm afraid you couldn't get very far like this. (In fact, strictly speaking, you couldn't get anywhere, per the example above. Processors themselves must store the instruction to be executed somewhere in order to process it) Both computers and humans require context for almost all non-trivial applications. If your word processor has no memory, where is your term paper? If you don't have any memory, how could you even spell a word, let alone form a coherent thought.

In the end, lack of memory may work if the entire problem can be solved in a single step, but that is about it.

Sorry.

 

[mpjesse]

So you're proposing that computers have some degree of natural intelligence?

The thing is artifacts and malfunctions are random by nature (take genetic diseases or retardation for example) Malfuctions cannot be totally predicted by an algorythm or a mathematical solution- even in computers. (Some might argue against this opinion based on a book called "A different kind of Science" written by Mathematica star Steven Wolfram. Steven belives there is in fact a mathmatical constant in nature that explains things like the patterns on a tiger and even life. I do not however, believe this to be true. Mathmatics are great for measuring and quantifying the universe, but they cannot begin to explain the complexity of it. And yes, I read his book so I can speak to it if anyone wants to argue!) Anyways, I argue that malfunctions or artifacts are random reactions to predictable outcome. But because these reactions (malfuctions in this case) are random this does not mean they are intelligent. (And until you prove otherwise, artifacts are indeed random side effects or malfuctions) DNA itself isn't intelligent; it's simply a set of instructions that have evolved and changed based on (among other things) environment conditions. Computers are at a very basic level no different from DNA. What you are trying to do is no different from what the effects on DNA might be if you tried the same thing- you are changing an environmental condition (in this case heat) to a device that has been programmed to operate within certain environmental conditions. We know something bad will happen, but what? The result is a random (though predictable) effect that results in artifacts or malfunctions. It is predictable in that we know malfuctions will appear if too high a temperature is reached, it is unpredictable in that we don't know exactly where how the reactions (or malfuctions if you will) will manifest themselves. Now if you apply this experiment to human DNA- you'd get the same results basic results as you would a computer. We know that if we expose an excessive amount of heat to a DNA strand- malfunctions will appear (if of course the strand was replicated and cloned). The result can be predicted on a very general level (like the clone would be unable to fuction correctly), but exactly HOW the clone would not be able to function would be a completely random effect. Take this rule and apply it to a computer. We know something bad will happen to the computer, but exactly what and what would the reactions be during the experiement? We cannot get the same result 100% of the time.

Perhaps this is a bad analogy to what you're trying to do. I think a better example would be applying direct heat to the brain. We know that the brain will eventually fry, shut down, and possibly die. However we cannot predict how exactly the person will react (emotions and behavior are in my opinion simply pre-programmed reactions that manifest themselves based on things like upbringing, experiences, age, sex, etc and hence have a random outcome). Certainly some people would die sooner than others in the same control settings. Others may scream out and cry while some may not make any noise. It is thus a completely random effect with a predictable outcome. Death. Just like your chip experiment.

I do believe that computers will one day become self aware and have the same type of natural intelligence that humans possess. The idea that computers are just machines and cannot ever have a conscence or "soul" is ridiculous. ("Souls" are nothing more than a belief in my opinion, so calling a thing "souless is ignorant anyways.) After all, what are humans? If you take "God" out of the equation, we are essentially no different than computers (on a very basic operating level of course). Additionally, computers are absolutely no different from ameboa's or protozoa or even viruses. Computers will one day no doubt be able to design organic existence for themselves (assuming of course this is an advantage to living orgranically). Since we're all made of the same basic elements, no one can say for sure which is supreme; bags of water and carbon that use chemicals to achieve intelligence or metal and silicon that use electricity to achieve intelligence. Both have their advantages and disadvantages; most of which are obvious so I won't go into it.

I think looking for intelligence in the form of malfunctions as a result of heat is the wrong place to start. True computer intelligence will almost certainly be purposely invented by humans and eventually evolve on its own. But do not let me discourage you. Contiune on with your quest.

If you are speaking of something else entirely, please ignore this reply.

-mpjesse

 

[koafc]

For memory, do you just mean RAM or do you memory of any type?

E.g. processors can store data in registers, L1 cache, L2 cache, L3 cache, RAM etc.

 

[stealth_blade]

Hey everyone. I don't know if anyone remembers my question a few months ago I posted on here about a theory that I have. It basically was that I believe that computers can exist without memory.

I want some input as to what I should do with my testing of this theory. I fully expect that if I do get any responses, I will be called stupid, go home kid, this is not a place for homework, idiotic, etc. But I hope that maybe atleast one reader will take me seriously.

I am putting my theory to the test currently that computers can indeed function without memory. What I am doing is getting very old computers and purposely overheating them with a heat gun. Why am I doing this? Well, I've seen video controllers overheat and produce artifacts on the screen. These artifacts are what you would consider a computer 'malfunction'. I propose that a computer malfunction, aside from software bugs, could possibly be the computer operating 'on its own'. A malfuction would be something that is not doing what it was intended to do.

Since computer circuitry that gets too hot is permanently damaged and will die prematurely, the key is finding the right balance of heat that can produce these artifacts without damaging the circuitry. This is proving to be very difficult, and since I don't know of anyone doing something like this, I don't know how to do it yet either.

I know that this all sounds crazy and a big waste of perfectly good computers, but I feel as though this could lead to something big. Please post some constructive criticism, and not just 'you are an idiot.' i get that enough each day.

I think a more basic question is in order here...how does overheating a cpu prove that a computer can operate without memory? The machine obviously had memory in it for you to turn it on and get it running in the first place did it not?

Secondly...A computer without any memory would be useless. It would be unable to process information because it couldn't store it. It would also be unable to return the results of any computation it was somehow able to make because that result would be instantly lost.

I think you also need to consider rewording your theory a little bit. You state "computers can exist without memory". Well, yes they can, but gasoline powered cars could exist without a gas tank. Neither would be useful however.

You also need to define the term 'memory'. At the most basic level, the cpu spends most of it's time moving data too and from various storage locations. Things such as harddisks, System Ram, CPU Cache, Registers, GPU RAM ect. All of these could be interpreted as 'memory'. So what do you mean by memory? Maybe you don't even know yourself?

 

[philderbeast]

a computer can definatly exsist witout memory there is no need to test that... but you cant do anything usefull with it.

the artifacts your refering to are a result of the silicon in the chip reaching a heat where the electrons are no longer controlled to moving throug the circut paths created but insted moving about ranbomly in the chip. so while you get a result (a garbled image) its not nessacarly the result of the chip working.

 

[oolceeoo]

When I mean memory, I mean ALL forms of memory. hdd, ram, rom, cache, registers, everything. A computer is comprised of input, processing, memory, and output. Take out memory from that list and you would get input, processing, and output. I'd venture to say that this would make a pretty powerful computer indeed!

Again, this is merely a THEORY, and I am not attempting to 'prove' that computers can exist without memory by applying heat.

Like mpjesse said, we all know what the end result will be if too much heat is applied: the computer will die. But what I'm trying to find out is what happens in the moments before the computer dies. At what temperature, or maybe possibly other conditions, can you 'observe' a computer on the verge of death.

The reason why I am starting to have a strong desire to pursue this seemingly mad quest is that I don't know what happens either. Neither does anyone that I've told my idea to. I've told my college professors about my idea, and even they do not know the answer.

I agree that yes, today's computers would be no use if they couldn't store data in memory. I'm not thinking in terms of current computing, and one person told me that possibly in the most likely VERY distant future that the paradigm of computing could drastically change into something chaotic and possibly indomitable.

 

[teamkiller]

u need to get out a bit more mate !

what are computers thinking nanoseconds b4 u fry them with a hairdryer, oh dear......

 

[Ned_Flanders]

You sir (and I really mean this) are an idiot.

 

[SU-37]

When I mean memory, I mean ALL forms of memory. hdd, ram, rom, cache, registers, everything. A computer is comprised of input, processing, memory, and output. Take out memory from that list and you would get input, processing, and output. I'd venture to say that this would make a pretty powerful computer indeed!

current processors need a program to tell the CPU what to do --->we need rom to store the program in .if we use a chip with the program just printed in it ...wow for every change we would make to the program we need a new chip -very economical indeed :lol: - its too costly and hard to maintain ,and the CPU will come like the old game consoles with program cartages. 8) .
by the way the computers today not input, processing, memory, and output. but input, processing, and output the memory and the processor is the heart of the mashine and contend as one ,all the other things are peripheries.
and your theory is not hard to make; by putting the memory in the input and the output devices. but it is a costly solution and will depend mostly on interfaces .
the future computers will have their memory in the processor itself and that what the scientists is working very hard on it today.

 

[ChipDeath]

A computer is comprised of input, processing, memory, and output. Take out memory from that list and you would get input, processing, and output.

Yes, but the input and output all reside in some form of storage. You're saying no form of memory, absolutely none. That means there's no physical way for any input to arrive at the processor anyway. You just end up with wires with a small current going through them - or lightbulbs, as they're otherwise known.

Directly input from a camera? Then the phsyical medium the camera is staring at becomes a form of storage. Same for a human inputting data - the human's brain becomes the storage device. Ad infinitum. YOU have to draw us a line somewhere, otherwise we're back to the lightbulb.

And as others have said, You NEED some form of memory to tell it what to do with the data. Add a small header to the data to describe what to do? Well It'll have to have some amount of memory somewhere on the chip to tell it what that header means... So That isn't going to work.

You cannot have a computer, in anything remotely resembling the current sense, without some storage medium.

And stop Bogarting. :tongue:

 

[RichPLS]

Can people live, walk and talk without a brain?
I think so, and see this every day.
Take the study of a chicken, who with a severed head ran around for days, and was fed by a tube.
Or a frog leg jolted with current jumps. Give a dead frog enough current in the right spots you could have it appear normal.
But that is not the frog on its own, it is something else causing this, and something else with a memory...
Take away the memory and what do you have, I forget.

 

[neocristi]

These artifacts are what you would consider a computer 'malfunction'. I propose that a computer malfunction, aside from software bugs, could possibly be the computer operating 'on its own'. A malfuction would be something that is not doing what it was intended to do.

This is your mistake: you watched the "PI" movie too many times probably and that is just science fiction kid.

The artifacts are just leaked currents as bits that flow in a wrong way in a chip working in stress conditions (maybe right before sommething melts in it). You don't want that to happen with your little Prescot core, although maybe it is happening and you don't know it :!: These Intel guys are very deceaving :roll: get an AMD

I'm going to propose you sommething REAL for a change, there is a scientist outthere that is building chips by connecting randomly millions of transistors, most of the chips do not work, but somme patterns were found that were impossible to think of! How do you think about that?

And by the way, forget about the missing memory, it must exist in one form or the other, otherway it is like talking about a storm on the moon !!!
See, there can't be wind without air (or somme sort of fluid).

Another thing, I'm impressed of how many people responded to this post, maybe everybody dreams of a super intellingent CPU that has self knowledge and is built by mistake (and costs $50 right?)

 

[Era]

When a transistor flips from one state to another and keeps that state as short or long time as needed, it is memory. So your quest after an apparatus without memory is a bit vague at least.
Even a broken light switch has a memory.

Maybe you are trying figure out how to make an Absolute Random Information Generator, or ARIG for short. The legends tell that the people of planet Mantsala in the eastern part of our galaxy ones tried that and managed to make one ARIG. They had invested their whole annual planetary budget in the project and they were very keen to sell it to somebody. Well, it turned out that nobody wanted anything as useless thing as an ARIG.
So they ran out of Coors and the whole humanoid population of planet Mantsala died in a global hangover and the pissants took over.

 

[RoundSparrow]

First off, what drugs are you on - as your thinking about the overheating is just 'out there'.

As someone already pointed out, a register within the cpu can be used as a form of memory - only 'one value per register' - but hopefully we can all agree that a register is part of the 'core cpu' and not like cache or other 'memory'. In other words, a register is a integral part of the cpu - and it is the simplest form of memory (one value held in cpu as long as the cpu has external electrical power).

A cpu can count and do lots of things with only registers... no external memory... so yes, a CPU can work fine without "memory".

It could even write to video cards, etc.

the real issue is that software pretty much is written to expect system memory (RAM)... and how to get the cpu to RUN your instructions... they tend to be stored in RAM.

But no reason you couldn't create a custom CPU and write microcode (on-cpu programming) that did something without RAM...

 

[endorphine44]

Take a few more chemistry and engineering courses. College professors are not always the smartest people on earth....more than a few are college professors because they can't make it in the real world market. You will apparently fit that category as well.
hopefully the extra classes will help you understand how things work.

 

[Finneus]

Well the only thing I can say about this thread is that I think its a commendable pursuit to try to think outside the box. I personally don't think what you are doing will work but sometimes the journey to enlightenment is cracking several eggs repeatly to understand the concept of a big friggin mess or the beginning of an omlet. I have a feeling this is indeed a huge friggin mess as has been stated repeated by people that I beleive have a much better handle on what is actually going on then I. But sometimes you only learn from the process....keep cracking those eggs maybe someday you will eventually end up with an omlet and not yolk over your face like this particular pursuit will probably end.....man I'm getting hungry. Good Luck!

 

[Rabidpeanut]

Okay, about the chicken, it still has nerves, it still has power, it just has no processor, ie a mobo can still run bios if no cpu is present.

Now those artifacts you are finding are just random crap, like stuff flying around in a vacuum. (Though there is nothing that is truly random).

But what you are attempting to do is use a static object to find intelligence, stasis precludes intelligence. Iow, if you find a computer that can fix itself then you can do what you are doing, but they are still only in development, and that is software-wise they still use memory. ALL inteligence has memory, ALL life has memory, DNA and RNA, so whatever you do you need memory, what you are doing, unless i am mistaken will not work. You need memory. Sorry to burst your bubble.

 

[RoundSparrow]

ok, he said WITHOUT REGISTERS.

Well... then I supposed you probably have millions of such "computers" around us all the time.

You are taking away a lot of what I would consider part of a "general purpose" or "programable computer". Without memory of some type to store the program to execute... then it becomes a set of "hard wired circuits" that does only a set of pre-defined tasks.

Can some hardware person help point this guy to the name of such "basic computing" such as "descreet logic" or other "non programable" "COMPUTING"?

 

[Zoron]

Without memory of some type to store the program to execute... then it becomes a set of "hard wired circuits" that does only a set of pre-defined tasks.

But it won't even do those "pre-defined tasks" without memory of some sort. Those tasks have to be stored somewhere in order to be executed... that means memory.

A computer won't function without memory... period. Even the most basic of computers (calculators and such) have some sort of memory. Even if a computer were intelligent... it would still require memory in order for it to perform any sort of function whatsoever.

So, a computer may exist without memory... but then it's not really a computer, as it wouldn't be able to perform any of the tasks that define a computer.

 

[Rabidpeanut]

Your brain has cells that act as registers (Actually synapses kinda do this) so to do what he wants will slow the computer like you can't believe. and then you will still need something like a register on the hdd.

 

[Era]

As someone already pointed out, a register within the cpu can be used as a form of memory

A form of memory? Register *is* memory!

but hopefully we can all agree that a register is part of the 'core cpu' and not like cache or other 'memory'.

Now you better define your concept of "memory".
Register no memory. RAM yes memory.

so yes, a CPU can work fine without "memory".

How?

It could even write to video cards, etc.

How? What would write there if it could? A purple dot?

the real issue is that software pretty much is written to expect system memory (RAM)... and how to get the cpu to RUN your instructions... they tend to be stored in RAM.

You don't understand much about computers or any other kind of logical devices. Do you?

 

[ara]

what if memory could be made from a calculation or the answer to a calculation is repeated over and over again while another processor reads the calculation... this means you would need a heck of a lot processors to 'make' memory where there is none, thus the system is able to 'function' without memory even though you have 'memory'

Ara

 

[Zoron]

The computer still needs instructions to perform the calculation... without memory, there is no where to store the instructions; therefore the calculation could not be performed.

Since a computer can be defined as a machine that performs calculations according to a set of instructions... you could NOT have a computer without memory. If you did, it wouldn't be a computer.