Sign in with
Sign up | Sign in
Your question

What is the cause of overheating?

Last response: in CPUs
Share
October 6, 2003 2:22:39 AM

My question is this, what is the main cause of all the heat generated by components? Be it in CPU's, Motherboards or whatever. Is it friction on a molecular level that causes the heat? Friction from all the electricity moving about, or just the electricity itself?

Just wondering

More about : overheating

October 6, 2003 3:44:03 AM

Electricity is hot!
October 6, 2003 4:23:48 AM

Electricity is not hot. It is the flow or electrons from one atom's valence ring to another's. It is the opposition to the flow that generates heat. Resistance is the oposition to current (electron flow). Part of this is friction on a sub atomic level, but also some is from collision.
Related resources
October 6, 2003 4:47:53 AM

Thank you!

So can I assume that if the resistance was eliminated it would not heat up?
Or even if it was reduced components would stay cooler, last longer and have a higher acceptance to overclocking?
Anonymous
a b à CPUs
October 6, 2003 6:51:17 AM

Captain Obvious says that lowering the temperature low enough would give rise to superconductivity, and resistance would be cut to zero!

But as transistors need resistance to work, your cpu would not function!

<b><font color=red>Captain Obvious To The Rescue!!!</font color=red></b>
October 6, 2003 7:25:22 AM

some heat is needed. if a component gets too hot or too cold, it won't function properly. also there are components that are called resistors. they take electricity and convert it to heat. they are mostly used like a regulator would be.

you can use gold for less resistance. its more dense and the atoms are closer together, but that gets expensive... or you can use diamonds... wouldn't get anywhere, but it'd be really expensive. anyway, when you are talking about semi-conductors, you need some resistance. the CPU core is of semi-conductor material (silicone)

</font color=red><font color=orange><b>visit my site. then send me your freakin picture.
<A HREF="http://www.lochel.com" target="_new"><font color=red>www.lochel.com&lt;/A>
October 6, 2003 3:16:41 PM

If we were to use gold, what would the results be?
Also, why does the processor need to be made from semiconductor material? Because of the needed resistance?
October 6, 2003 4:13:41 PM

Quote:
What is the cause of overheating?

Too much heat!

Sorry. :\ I couldn't help myself. Everytime I read that title I just wanted to yell that out. (Just like every time at the movie theater when they throw out the slogan "What's a movie without a tasty snack?" I want to yell out "A MOVIE!") The blatantly obvious is just so entertaining. :) 

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>
October 6, 2003 4:18:17 PM

On a serious note I thought that it was the electron migration that caused most of the heat. I mean ideally the juice should always flow along the path of least resistance (AKA the paths that were etched for it) but in reality a lot ends up bleeding off in useless directions, thus generating heat. Can anyone please correct me on that if I'm wrong? :) 

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>
a b à CPUs
October 7, 2003 8:09:35 AM

Transistors are made of semiconductors because of the way diods work.

A diod is a material that flows electricity with low resistance in one direction and a high resistance in the other. Ideally you'd want perfect conduction in one direction and no conduction in the other, but there aren't any materials that can do it.


<font color=blue>Only a place as big as the internet could be home to a hero as big as Crashman!</font color=blue>
<font color=red>Only a place as big as the internet could be home to an ego as large as Crashman's!</font color=red>
October 7, 2003 1:14:08 PM

So the electronics work on a principle of resistance, and resistance causes heat, so the electronics are (more or less) designed to generate heat. Obviously that's not their <i>main</i> purpose, but still...

Man I can't wait until electronics are made from things like carbon nanotubes. :) 

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>
October 7, 2003 3:28:37 PM

Thanks again guys!
I think I understand now?!
a b à CPUs
October 7, 2003 9:08:30 PM

heh, that's a good way to turn things around, but in reality they're designed NOT to produce heat. A perfect material would produce no heat in the direction of current flow (perfect conduction) and no heat from leakage of current flowing backwards (perfect isolation).

<font color=blue>Only a place as big as the internet could be home to a hero as big as Crashman!</font color=blue>
<font color=red>Only a place as big as the internet could be home to an ego as large as Crashman's!</font color=red>
October 7, 2003 10:42:40 PM

i just saw this on daily planet last nite.

sometihng this guy calls fire paste.

Can a material discovered by an inventor in North Bay, Ontario be the next great advance in heat-resistent material? <A HREF="http://www.exn.ca/video/?video=exn20030903-firepaste.as..." target="_new">Watch it here</A>

ignor the first minute... its the next 3 minuts you wana see.

Crazy stuff! Not even Nasa came up with this stuff... and they have a freeking team of scientists!!! HA

anyhoo...maybe in years to come they will figure out a way to take this stuff down to its smallest melecualar size..who knows.



ASUS P4S8X - P4 2.4B - 2 x 512M DDR333 - ATI 9500 Pro(Sapphire) - WD 80G HD (8M Buffer) - SAMSUNG SV0844D 8G HD - LG 16X DVD - Yamaha F1 CDRW - Iomega Zip 250 int.
October 8, 2003 4:13:35 PM

Quote:
heh, that's a good way to turn things around, but in reality they're designed NOT to produce heat. A perfect material would produce no heat in the direction of current flow (perfect conduction) and no heat from leakage of current flowing backwards (perfect isolation).

Agreed, but in the lack of a perfect world with the perfect material...

Plus it sounds like the 0.09micron process is really proving the fact that they <i>aren't</i> using the perfect materials for electronics. Heh heh. I hope that they get that sorted out.

I have to wonder though, what <i>is</i> the best electrical insulator that we have for a scale that small?

And I also have to wonder if the human nervous system contributes to our body heat for the same reasons that electronics get hot. :\ Something in the back of my faulty organic storage medium is saying that it's based on a chemical reaction, not an electrical reaction, so it doesn't contribute heat. But then another part of that faulty medium is recalling that chemical reactions and energy changes go hand in hand too. My poor faulty organic processor is all confused.

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>
October 9, 2003 5:51:36 AM

ok so a processor is vertually all switches,(which cause amajority of the heat) do they require an eletrical current...or is light concievibly possible?

I should probalby stop here but...

i would guess that the CPU would need to be far bigger, meaning the switches would be further apart...slowing things down some...light travels faster and may make up for this...(i have no idea)

but light has an edge(not just being cooler), electrical current only has frequentsies(i can be wrong, enlighten me if there are other ways), but light can have frequentsies, hues, and shades.

this could change the whole process completly removing the switch from being only on/off or I/O to having multiple reactions.

all this with little to no heat.

is this concievable? or am i going WAY off base here?

ASUS P4S8X - P4 2.4B - 2 x 512M DDR333 - ATI 9500 Pro(Sapphire) - WD 80G HD (8M Buffer) - SAMSUNG SV0844D 8G HD - LG 16X DVD - Yamaha F1 CDRW - Iomega Zip 250 int.
October 9, 2003 3:48:51 PM

Quote:
ok so a processor is vertually all switches,(which cause amajority of the heat) do they require an eletrical current...or is light concievibly possible?

Some have tried to develop optical CPUs. None have publicly gotten anywhere. To use light you need something that generates light, and for that you need a HUGE CPU.

Quote:
i would guess that the CPU would need to be far bigger, meaning the switches would be further apart...slowing things down some...light travels faster and may make up for this...(i have no idea)

Yes, they'd be MUCH bigger. Electricity travels pretty fast as it is, and across the increasingly small distances on CPUs I would dare say that electricity is a better medium than light. The distances involved for an optical CPU to work at present are just too great.

Optical components would work well in other areas of the CPU though. For example to have an optical cable connecting hard drives would be useful. Replacing wired audio cables with optical audio cables would be nice too.

Quote:
but light has an edge(not just being cooler), electrical current only has frequentsies(i can be wrong, enlighten me if there are other ways), but light can have frequentsies, hues, and shades.

Actually, the frequency of the electro-magnetic radiation determines the hues and shades of light, so light still only has frequencies. However the electrical systems don't work off of frequency calculations but simply of on-off gates (bits). So any CPU which could handle calculations per switch using several frequencies of light would have an interesting advantage there by having bit-states that are more than just 0 or 1.

The incredible flaw however is that a light frequency-based CPU would require a storage medium capable of either storing bits that hold values of more than just 0 and 1 (which is no easy task) or else of using up tons of space in traditional storage mediums.

For example, say that you have an optical CPU that uses color frequencies to have 256 unique values (0-255) per switch (a 256-value bit). Now say that you loaded up that system with 1GB of RAM using the traditional binary bits. That would mean that your 1GB in binary is now just 128MB in color light.

Quote:
this could change the whole process completly removing the switch from being only on/off or I/O to having multiple reactions.

Yes, it could. But until storage devices are capable of storing those multiple reactions it will mean a very big mess.

Quote:
all this with little to no heat.

I'm not so sure about that. Lasers burn for a reason. Light is not without it's own heat-generating principles.

Quote:
is this concievable? or am i going WAY off base here?

I'd say about half and half. None of your ideas are new, but no one has as of yet figured out how to actually make them possible either. I'd dare say that we're a lot closer to computers that process information based on the energy states of electrons in an atom than we are to computers that process information based on colored light.

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>
October 9, 2003 4:34:58 PM

Electrons (or electricity as it was stated) move at the speed of light (186,000 miles per second). The devices previously referenced to as "diods" are actually diodes. A basic PN-junction device. Heat is caused by current flow through resistance. Resistance is used to control the current flow for various applications. A CPU is basically a bunch of transistors working off one another and an electrical current is necessary to bias the junctions inside the transistor.

The idea of computers processing information based on colored light seems far fetched. I'm not quite clear on slvr_phoenix's idea about frequencies and switches. Switches work on the principle of having a signal there or not regardless of it's frequency. You could have a multi-purpose switch that responds to many frequencies and performs various actions but it would still be based on that frequency being there or not.
October 9, 2003 5:17:45 PM

very well detailed responce, thanks! many very good points. spose i shoula know that these ideas have been thought of already, i guess i was just currious as to how close we are to it, or how plausable it is to do.

interseting that you would point out that the RAM would drop from 1GB to 128MG. never realy thought about that... but at the same rate, when going to say 256 bit values the CPU's bit rate im sure would be changed as well rather then the 32 or 64 bit, to something more like 1024(not sure what the conversion there would be)... making RAM virually unlimited...no?

That would make one hell of a supercomputer! even it is is the size of my desk...its computational power would be massive...even if it were slow. the amount of information it could computer in one pass would be astronomical...unfortunatly that would be all it'd be good for... Massive computation. i would think, for things like space and perhaps medical perposses.


remids me of a project i had while in college. see if i can dig that up. discribing the differance between 32bit and 64 bit processing. i related it to rollercoster rides.


ASUS P4S8X - P4 2.4B - 2 x 512M DDR333 - ATI 9500 Pro(Sapphire) - WD 80G HD (8M Buffer) - SAMSUNG SV0844D 8G HD - LG 16X DVD - Yamaha F1 CDRW - Iomega Zip 250 int.
October 9, 2003 6:05:59 PM

Quote:
The idea of computers processing information based on colored light seems far fetched. I'm not quite clear on slvr_phoenix's idea about frequencies and switches. Switches work on the principle of having a signal there or not regardless of it's frequency. You could have a multi-purpose switch that responds to many frequencies and performs various actions but it would still be based on that frequency being there or not.

Our electrical switches work on that principle, yes.

Imagine however a 'switch' where the present state is defined a one frequency of light and the state change is defined as a second frequency of light. The resulting frequency of light would be a third value. You could literally use this method for a single 'switch' to instantly sum a great number (the number depends on accuracy of the hardware) of unique values. An optical CPU (if designed right) is very different from binary logic.

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>
October 9, 2003 6:13:49 PM

Lol d0rk when did you become smarter than Einstein, electrons at speed of light then eh? Well let's think about it at a basic level if they did travel at that kind of speed you would have a nice little blackhole of infinite density inside your PC case now wouldn't that be fun? Since electrons have a mass (a minimal one of 9.1 x 10^-31 kg) that mass starts to increase as they start to travel closer and closer to the speed of light, so theoretically nothing that consists of any matter will ever reach the speed of light. The actual speed of electrons is pretty close to that most of the time. The speed is proportional to its energy level given by the equation 1.602 x 10^-19 Joules for one electron losing one volt of potential difference (eV) Sorry to be nagging I don't mean to be scornful, but this is basic physics get yer facts straight :) 
October 9, 2003 6:14:55 PM

Quote:
interseting that you would point out that the RAM would drop from 1GB to 128MG. never realy thought about that... but at the same rate, when going to say 256 bit values the CPU's bit rate im sure would be changed as well rather then the 32 or 64 bit, to something more like 1024(not sure what the conversion there would be)... making RAM virually unlimited...no?

No. The number of values that a 'bit' represents is used equally both ways. So the RAM in the end has the exact same potential of 1GB binary even if it's 128MB with 256-values per bit. In theory the amount of storage remains the same.

In practive however that isn't necessarily true. In a binary system you have a wider range of variable sizes than you do in a 256-values per bit system. That flexability means that you'll be better able to utilize the space that is available because you won't be forced into using storage capacity for numbers that are much larger than what you need.

This is actually also where lazy porting to 64-bit software can be trouble because if you just offhandedly converted all 32-bit integers to 64-bit integers you'll double the storage space needed for holding the exact same information. The same thing happened back when we went from 16-bit to 32-bit processors. Lazy programmers deserve to be smacked. :o 

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>
October 9, 2003 7:25:58 PM

In theory electrons move at the speed of light. I wasn't going to go into physics for a simple explanation. And if you want to get real in depth, the speed of light can be varied, so it could actually be possible that electrons move faster than light. And to answer your question, yes a nice little blackhole of infinite density would be fun, how many could claim that honor? I could slap an "Infinite blackhole of density inside." sticker on my case right next to the AMD one :) 
October 10, 2003 5:06:43 AM

blackhole of infinite density inside eh...that explane's alot...wish i coulda used that one on my teachers

ASUS P4S8X - P4 2.4B - 2 x 512M DDR333 - ATI 9500 Pro(Sapphire) - WD 80G HD (8M Buffer) - SAMSUNG SV0844D 8G HD - LG 16X DVD - Yamaha F1 CDRW - Iomega Zip 250 int.
!