Self-Cooling Observed in Graphene

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
Researchers from the University of Illinois have discovered that graphene transistors have a self-cooling effect that could lead to graphene-based chips that require little or no cooling.

"In silicon and most materials, the electronic heating is much larger than the
self-cooling," King said. "However, we found that in these graphene transistors, there are regions where the thermoelectric cooling can be larger than the resistive heating, which allows these devices to cool themselves. This self-cooling has not previously been seen for graphene devices."
 
This would be good news indeed if it can be backed up/confirmed by independant sources repeating the research. Otherwise, I'm going to take this a bit of Cold-Fusion. The idea of self-cooling electronics that never warm up as long as you don't take them past a certain voltage sounds wonderful but unrealistic. Law of conservation of energy and this don't seem to mix only b/c I dont see how the energy is leaving? Total Energy in System = Energy already in system + energy inc to system - energy leaving system. If 400 watts of energy is going in, well, you'd expect the total energy in the system to increase unless 400 watts was also leaving. I suppose it could be disapaited into the air, but, you'd still need case fans to remove the heat from the air.
 
You DO know that they did discover cold fusion back in the 70s, but then they buried the research lab for the fear that the reds were going to steal it. Just saying....
 
Sounds like this a substance that heats up fast, and cools down fast as well, which makes cooling pretty easy.

... I dont see how the energy is leaving? ...

From the arcticle:
... The measurements revealed surprising temperature phenomena at the points where the graphene transistor touches the metal connections. ...

It's transferring the heat to the connections. I'm guessing that a non-conducting, finned metal 'top' might be good enough to keep it cool.
 
This would be good news indeed if it can be backed up/confirmed by independant sources repeating the research. Otherwise, I'm going to take this a bit of Cold-Fusion. The idea of self-cooling electronics that never warm up as long as you don't take them past a certain voltage sounds wonderful but unrealistic. Law of conservation of energy and this don't seem to mix only b/c I dont see how the energy is leaving? Total Energy in System = Energy already in system + energy inc to system - energy leaving system. If 400 watts of energy is going in, well, you'd expect the total energy in the system to increase unless 400 watts was also leaving. I suppose it could be disapaited into the air, but, you'd still need case fans to remove the heat from the air.

actually your definition of energy seems to be heat only but energy can be in other form
say you have 1KJ going into your system but it's so efficient that there's almost no resistance in the system then most of the energy will go into actually doing stuff
what i don't understand is that i thought in order for something to get colder something else has to get warmer so... what's getting warmer?
 
Sounds like this a substance that heats up fast, and cools down fast as well, which makes cooling pretty easy.

It's transferring the heat to the connections. I'm guessing that a non-conducting, finned metal 'top' might be good enough to keep it cool.

Great so the 'heat' goes into the metal touching the graphene....now where does the heat go from there? Obviously, that heat still needs to be removed from the 'system' that is your tower/case. That seems akin to saying 'Heat from CPU where touching CPU Heatsink goes into CPU Cooler! Great Success! No fans/water cooling needed anymore!' except the glaring obvious problem that the cpu-cooler unless passive requires a fan and you still need case fans.
 
actually your definition of energy seems to be heat only but energy can be in other form
say you have 1KJ going into your system but it's so efficient that there's almost no resistance in the system then most of the energy will go into actually doing stuff
what i don't understand is that i thought in order for something to get colder something else has to get warmer so... what's getting warmer?

Yes, with most electrics, heat is the major source of energy transfer. I didn't really bother to touch transfer via sound or light. Mostly, b/c if your hearing popping and crackling sounds from your gpu/cpu/pc's transistors, your screwed.

To an extent, you might see energy transfer in the form of 'light' such the tsugen in a lightbulb//an led but generally if you have a cpu/gpu chip that's become so hot, it's turned red and is being to emit light, .... your cpu/gpu is melted/screwed beyond belief.

From a link in the article above, the energy seems to be going into the metal connector but at the same time, that heat still needs to be disipated from the metal connector.
 
well no it still needs energy to flip transistors from one state to another state so if your stuff is 100% efficient then all the energy being used will be spent doing that and none to heat
i thought the article said the point where the metal connector is at is the cooling zone meaning it's not generating excess heat and is cooler than normal
 
Seems the more they develop and research graphene resistor tech., the more promising it becomes.
 
Kind of neat but...

The Illinois team used an atomic force microscope tip as a temperature probe to make the first nanometer-scale temperature measurements of a working graphene transistor.
Materials do really freaky stuff at tiny scales, if this material did this at a macroscopic scale it'd be much more impressive.

It will be interesting to see where this goes though.
 
It's to bad, not much will come out of graphene research if anything at all other than general observations. Good luck commercializing any graphene product.
 
For the big bunch that didn't get it, it isn't that they get cold when used, it is that they generate very little heat and are great at dissipating the heat generated.

If able to be used en masse then it would be awesome for ultra portable devices, and in general it could give quite a nice speed bump, allowing you to run a chip at higher speed than a silicon based alternative, and both using the same cooler.
 
Why no commercialization?

I never said no commercialization, I just said, good luck in commercializing it. Graphene will be difficult to commercialize due to it's fragile nature. IBM even admits it can't replace silicon, but they may find a way to hybridize it. It's going to be a tough thing to do.
 
all i want to know are two things:

1. will this actually come to light so that i can have a cooler cpu?

2. how much harder can i overclock that bitch?
 
Self-cooling CPU chips, whoa.

Watercooling will still be a lot cooler, no? I doubt it will go anywhere soon, but this is a huge help to people as its just one less step in building a PC. One of the more pain in the ass steps too :)
 
Seems the more they develop and research graphene resistor tech., the more promising it becomes.

And this is a good thing. All we need now are room temperature superconducting materials and we're set.
 
Back
Top