i do a bit of overclocking in my spare time but nothing to extreme i know that you need a good cooling system.
But i also know that some people use liquid nitrogen to OC.
But as far as i know silicon being a semi-conductor as the temperature gets lower its resistance increases...
I can understand the need to keep the temp under the melting point of silicon but whats the point in using LN surrely that would diminish its OCing ability...
Jules
overclocking
Re: overclocking
It is the other way round... Electrical Resistance drops at lower temperatures, it is one of the reason's (or possibly should be explained the other way round) that lightbulb's work.
try http://en.wikipedia.org/wiki/Semi-conductor
look at Carrier generation and recombination
quote: "the product is a function of the temperature, as the probability of getting enough thermal energy to produce a pair increases with temperature, being approximately 1/exp(band gap / kT), where k is Boltzmann's constant and T is absolute temperature."
This because in semi-conductors the conductivity is less effected by Brownian motion than by the release of electrons into the molecular lattice as opposed to metals.
I know this for a fact because i know a guy how tried to use liquid helium (4.22 K, -268.93 °C or -452.07 °F) to coll a p4 except for it destroying the gold connections the die of the processor actually stopped conducting.
So i still don't understand why liquid nitrogen cooling helps OCing so much except that it completely eliminates any thermal dissipation problems
Jules
look at Carrier generation and recombination
quote: "the product is a function of the temperature, as the probability of getting enough thermal energy to produce a pair increases with temperature, being approximately 1/exp(band gap / kT), where k is Boltzmann's constant and T is absolute temperature."
This because in semi-conductors the conductivity is less effected by Brownian motion than by the release of electrons into the molecular lattice as opposed to metals.
I know this for a fact because i know a guy how tried to use liquid helium (4.22 K, -268.93 °C or -452.07 °F) to coll a p4 except for it destroying the gold connections the die of the processor actually stopped conducting.
So i still don't understand why liquid nitrogen cooling helps OCing so much except that it completely eliminates any thermal dissipation problems
Jules
-
piranha
- Member
- Posts: 1391
- Joined: Thu Dec 21, 2006 7:42 pm
- Location: Unknown. Momentum is pretty certain, however.
- Contact:
I have a few question about over clocking...
My bios settings have only Enable/Disable for overclocking.
Does it decrease the stability of the OS?
Is it dangerous to the processor?
Does it significantly increase the speed of the computer?
Thx
-JL
My bios settings have only Enable/Disable for overclocking.
Does it decrease the stability of the OS?
Is it dangerous to the processor?
Does it significantly increase the speed of the computer?
Thx
-JL
SeaOS: Adding VT-x, networking, and ARM support
dbittman on IRC, @danielbittman on twitter
https://dbittman.github.io
dbittman on IRC, @danielbittman on twitter
https://dbittman.github.io
Firstly i would like to say i am sure that CPU dies do not conducting with liquid helium i will look into the conductivity issue at higher temps and can you give me the url of your source that says the opposite please.
To piranha:
firstly you must now that when you overclock you loose you're CPU's warranty is void and possibly your whole systems warranty depending on if you made your system your self or bought.
if your bios only gives you a enable/disable it could mean two things or your bios has a fixed overclock (e.g it increases your bus speed by 2%) or once you select this option it gives you more options to overclock.
it can decrease the stability of your OS but this depends on how much you OC. If you have stability problems you can also increase the CPU voltage but i would not recommend increasing it more than 0.25V above default.
is it dangerous to the proc??
It depends on you're colling system what CPU you're using and obviously how much you OC it. right now im running a C2D E6600 (default 2.4GHz) at 3.4GHz with a voltage increase of 0.275V (I would not normally do this but i'm going through a performance freak phase..., I would not recommend pushing a C2D above about 3.2GHz for normal use).
Is it dangerous to the proc??
again it depends how much you increase the frequency/voltage and on your cooling system.
Also remember that not all CPUs come out of the fabs with the same abilities some will be able to overclock lots and some not so much.
The trick is frequent stability test and not going crazy on increases (i.e. small increases at a time).
Does it significantly increase the performance of your system???
that depends on to things:
1) obviously how much you overclock it 10MHz is not going to make a visible difference...
2) on the rest of you're hardware if it cant use the extra power then you're just going to end up having a CPU that is latent most of the time.
As an example of performance increase with my CPU running at normal speeds I get a 3Dmark06 score of 9876 when its overclocked i get 10792 (ok i did overclock the GPU a bit but not much)
I also asume you now how youre bus speed affects ure CPU speed and frequency multipliers and all that but if you don't hesitate to ask.
You can reach me at [email protected]
Jules
To piranha:
firstly you must now that when you overclock you loose you're CPU's warranty is void and possibly your whole systems warranty depending on if you made your system your self or bought.
if your bios only gives you a enable/disable it could mean two things or your bios has a fixed overclock (e.g it increases your bus speed by 2%) or once you select this option it gives you more options to overclock.
it can decrease the stability of your OS but this depends on how much you OC. If you have stability problems you can also increase the CPU voltage but i would not recommend increasing it more than 0.25V above default.
is it dangerous to the proc??
It depends on you're colling system what CPU you're using and obviously how much you OC it. right now im running a C2D E6600 (default 2.4GHz) at 3.4GHz with a voltage increase of 0.275V (I would not normally do this but i'm going through a performance freak phase..., I would not recommend pushing a C2D above about 3.2GHz for normal use).
Is it dangerous to the proc??
again it depends how much you increase the frequency/voltage and on your cooling system.
Also remember that not all CPUs come out of the fabs with the same abilities some will be able to overclock lots and some not so much.
The trick is frequent stability test and not going crazy on increases (i.e. small increases at a time).
Does it significantly increase the performance of your system???
that depends on to things:
1) obviously how much you overclock it 10MHz is not going to make a visible difference...
2) on the rest of you're hardware if it cant use the extra power then you're just going to end up having a CPU that is latent most of the time.
As an example of performance increase with my CPU running at normal speeds I get a 3Dmark06 score of 9876 when its overclocked i get 10792 (ok i did overclock the GPU a bit but not much)
I also asume you now how youre bus speed affects ure CPU speed and frequency multipliers and all that but if you don't hesitate to ask.
You can reach me at [email protected]
Jules
Oh and another thing your BIOS should stop your system booting if you try anything to mad so i would not overclock if your mobo doesn't have a reset CMOS button and it also does diminish the life of your proc but if you don't go to mad then by the time it fails you will probably have replaced your system by then (i.e the decrease in lifespan for overclocks that don't push your CPU to the limit wont be strongly diminished), but again remember you should look at results that other people have had for overlclocks of the same CPU model so you no about were to aim and again not all CPUs of the same model have the same overclock ability.
Jules
Jules
Re: overclocking
Tyler, I was stunned to find that he is absolutely right: Conductors increase resistance with temperature, semiconductors have it the other way 'round.Tyler wrote:It is the other way round... Electrical Resistance drops at lower temperatures...
Every good solution is obvious once you've found it.
Think of it like this: No matter the conductivity of the semiconductor, teh CPU will always produce heat. The higher the voltage / clock rate, the more heat is generated. The better the cooling assembly, the better the heat is gotten away from the core.
Bottom line, the better the cooling, the higher the voltage / clock possible before the core dies a violent death.
PS: I consider OC'ing to be a waste of time, energy, and ressources. I value silence (and, for laptops, battery life) over raw performance. The next chip generation will beat your OC'ed system anyway, but silence is eternal.
Bottom line, the better the cooling, the higher the voltage / clock possible before the core dies a violent death.
PS: I consider OC'ing to be a waste of time, energy, and ressources. I value silence (and, for laptops, battery life) over raw performance. The next chip generation will beat your OC'ed system anyway, but silence is eternal.
Every good solution is obvious once you've found it.
Re: overclocking
Way to destroy my Universe...Solar wrote:Tyler, I was stunned to find that he is absolutely right: Conductors increase resistance with temperature, semiconductors have it the other way 'round.Tyler wrote:It is the other way round... Electrical Resistance drops at lower temperatures...