The next big thing in computer cooling

N

NeoNemesis

2[H]4U
Joined
Mar 10, 2004
Messages
2,467
As computer components get faster they also get hotter and hotter. This means that baring excessively sound fans, air (and water) won't be able to provide cooling for components in 10 or so years. My guess is that liquid nitrogen or liquid hydrogen will eventually be used to replace them altogether.
 
thank you for your newbie cheez I can see into the future too and I dont see computers anymore I see a whole destop in my head with audio output;) sorry you asked for it
 
air=easy bc it exists in our rooms
water=pretty easy bc it is easy to dissipate heat to air and recycle back into system (ie, radiator, etc)

how do you "foresee" easily recycling LH or LN2 so that it can actually be reused?

it's not like you can just pass it through a radiator...
 
Thanks for the enlightening feedback. I was just suggesting a possibilty but if you... oh great limp gawd... have a different opinion of what cooling will be like in 10 or 15 years then I humbly request that you share it. :)

Can't people even speculate about things that would be cool have in the future without comments like the ones above?

I'm not a chemist and this is not the future so I really don't know how you could do it, but since there is an abundance of hydrogen and nitrogen on this planet, I can only assume that eventually, the technology to do this will be invented.
 
You do realize that computer hardware in "10 or so years" will be radically different from computer hardware today and those components, most likely, will be designed to run at either slightly hotter, slightly cooler, or the same temps as current hardware. Both of the cooling methods which you listed are very poor long-term cooling solutions due to their extremely cold temperatures (which would introduce a lot of condensation into the computer) and their tendency to vaporize at room temperature. To contain either of those substances into a 'lossless' cooling solution would basically be the same as creating a "super Prommie" of sorts. The fact of the matter is that no cooling solution, that I recall, can match the performance to price ratio as well as the sheer simplicity and reliability of air cooling--which is why it is the most common form of cooling in personal computers today. (Not to mention that air cooling, once setup, is pretty much idiot proof)

I don't think there will be much change going on... We'll prolly see a lot more water cooling going on in the next few years but then, as newer components become more efficient and put out less heat, computer manufacturers will drift back into air cooling.

Then again, we may both be totally wrong. :rolleyes:
 
You are probably right, liquid hydrogen has its uses (cooling fission reactors, powering fuel cell cars) but it probably won't ever be used for cooling in pc's. I think it will be interesting to see how computers will change in the next decade.
 
Originally posted by NeoNemesis
Thanks for the enlightening feedback. I was just suggesting a possibilty but if you... oh great limp gawd... have a different opinion of what cooling will be like in 10 or 15 years then I humbly request that you share it. :)
Click to expand...
I'll share my opinion.

I think:

1. Threads like this, (speculation based), are pointless.
2. You forgot something: Advances in technology will bring about hardware with no cooling requirements.
3. Nobody cares if you are a 'newbie cheeze' if your post content is good. Your response to Kenny's childish argument was even more childish.
4. Rise above the cheese status. You can do it.
5. ...
6. Profit.
 
One would hope that the brains of the world would build hardware that performs better at lower power consumptions levels...therefore produce lses heat = no need for extreme cooling solutions
 
It's been tried both successfully and unsuccessfully. The successful approach used liquid nitrogen immersion of the CPU. He was able to double the speed of an Intel CPU from 2 GHz to 4 GHz. The previous attempt was from a fellow who tried to use a 3M fluorocarbon fluid for cooling the CPU and tried to use liquid nitrogen to chill the "cooling fluid". Since the cooling fluid was a liquid at room temperature, it promptly froze when chilled to77° Kelvin.
Click to expand...

from http://www.7volts.com/cooling.htm

pics at http://www16.big.or.jp/~bunnywk/ln2/ln2_e.html
 
Liquid hydrogen would be the most devastating cooling technology to implement, any computer user would risk serious explosion with all the gaseous hydrogen in the air (if you recall, hydrogen is extremely explosive ... not really a good thing to have sitting on your desk). As for Nitrogen and Hydrogen being 'abundant in the atmosphere', maybe in gaseous form, but to produce liquid nitrogen or hydrogen requires excessive amounts of energy and resources, to even think of using these as widespread cooling techniques is unbelievable, it would cost you some SERIOUS money just to cool your PC.

As far as speculation goes, technology is so diverse and changing its really hard to say. There are a number of R&D resources being funneled into research of quantum chip design, using other organic substrates to replace silicon, photon based (rather then electron) optical chips, etc. Since none really exist yet (at least none capable of being commercially manufactured) it is hard to say if they ever will, or if they will solve the heat crisis, either way, 10 years is a long time, and a lot of things can and will happen in that time. As the saying goes "Necessity is the mother of invention".
 
I think in the next few years you will see more and more instances of combination liquid/air solutions that will use heatpipe like systems to move heat from hot components to air chambers that optimize airflow to improve cooling efficiencies. The Intel BTX case is an attempt to move components arouns but I think it can be done in a much more efficient way. Think of a wind tunnel with a radiator that will cool a liquid that is connected to hot components via some type of passive pump. Or small highly reliable electric pumps to move some thermal fluid. The goal would be channel as much external cool air over the largest possible surface area while channeling exhaust air as far away from the intake as possible. some type of fan would be used to draw air threw the channel - it would basically look like a jet turbine from a cross section view. I am not a chemist but I am sure that there are fluid options that could be used that are more efficient than water and I have read a few articles on passive pumps and miniature pumps that could be used.
 
The future technology will be focused on processing technology, and not the cooling. Once we make a few things, we'll be back to passive cooling, if any at all.
 
the via cpus are pasivly cooled arn't they. future is going to produce smaller more efficent cpus unlike the prescott:rolleyes:
 
Seeing how in 10 years accoring to Moores Law we will need new materials in Transistors, there is no definate answer to your idea or question right now.

I don't know about AMD, Cyrix, Transmeta, IBM, Motorola, or any other chip makers, but so far, the only company to design a CPU that needs phase change cooling for a CPU is Intel. The Itanium2 comes to mind. It might be slightly dated, but it generates enough heat to heat my room in the dead of winter. The Prescott core also generates a whopping amount of heat. Until some new technologies get finished, cooling needs will stay about even. With a few new technologies, cooling requirements might actually relax a little. Right now isn't the best of times to be guessing about it...

So far, the most effective form of cooling is water. Water transfers heat four times faster than air. You still have to transfer the heat from the water to air, but that is easier than going from the CPU core to aluminum to air.
 
In the future we will have nanocomputers or computers that use biological chips and stuff like that that won´t produce any heat at all :D

Believe me you won´t believe what computers it´s out in 10 years.
 
hey back in the day people thought that we would be in space ships by now flying around like the jetsons. don't get too far ahead;)
 
Originally posted by NeoNemesis
Thanks for the enlightening feedback. I was just suggesting a possibilty but if you... oh great limp gawd... have a different opinion of what cooling will be like in 10 or 15 years then I humbly request that you share it. :)

Can't people even speculate about things that would be cool have in the future without comments like the ones above?

I'm not a chemist and this is not the future so I really don't know how you could do it, but since there is an abundance of hydrogen and nitrogen on this planet, I can only assume that eventually, the technology to do this will be invented.
Click to expand...


lol when you make it sound so definate you can't blame people for getting pissed at your sheer stupidity.

To speculate though, I doubt we will have todays cooling issues in 15 years. Most likely computers will continue to become smaller and smaller and require less cooling facilities to operate, if it became any other way we would probably not have the means to keep our systems cool enough to keep increasing their speeds. Just look at new technologies like DDR2 and BTX they all seek to combat temperatures in a way to allow faster computers in the end.
 
Originally posted by thebro
hey back in the day people thought that we would be in space ships by now flying around like the jetsons. don't get too far ahead;)
Click to expand...

We are. If cars had developed in the same rate as computers we would everyone fly to the moon with our cars as of today.

If we could translate the computer technology to cars we would be able to do just that.
 
Originally posted by ’m‚³‚ñ
So far, the most effective form of cooling is water. Water transfers heat four times faster than air. You still have to transfer the heat from the water to air, but that is easier than going from the CPU core to aluminum to air.
Click to expand...
How do you figure that... there is much better cooling around than water, albiet much more expensive and difficult.
 
I don't know what somepeople think about PC's but PC and Intel P4, and AMD XP's are x86 compatible computers...

If you gave up on that system and moved on to a different scope of instructions with eg 128 bit or 256 bit cpu's perhap even 1kb wide cpu's you'll have to use FARR less cpu cycles to accomplish the same task.
When a cpu has a reduced cpu clock speed it doesnt have to push and pull that many electrons for every 'stroke' == Less thermal loss == cooler.

Digital Signaling Processors are often designed from the ground up, and ive seen a DSP under 10 watt that delivered 4000 MIPS!


(Ofcourse 1kb wide cpu is a bit uncomfortable, but we might see that in the future...)

me boring? :rolleyes:
 
In the future if you start up a pc and the fan goes down on the heatsink they will be so hot you'll experience nuclear meltdown.

or maybe they will make chips that are more efficent and make less heat.

but I won't double the first idea.
 
How do you figure that... there is much better cooling around than water, albiet much more expensive and difficult.
Click to expand...

Thats why water is the most EFFECTIVE cooling option, because the cost per wattage of heat dissipated is the lowest.
 
I just have one fear..if computers become too small so they can fit into ones palm, what will happen to case modding? a whole hobby lost for many here:(


:D
 
I imagine people in the future once they have powerful enough computers will make programs that will make the code in their existing programs more efficient. If you wrote your whole video encoding program in ultra efficient ASM ( which would take FOREVER for a human, hence the computers could be the optimizers ), it could possibly encode twice as fast on the same hardware. I think that the emphasis on fast code will become an issue unless people think up some way to make processors move information faster than the speed of light, which moores law will probably require in the next 20 years.
 
Originally posted by HiTech-Hate
Thats why water is the most EFFECTIVE cooling option, because the cost per wattage of heat dissipated is the lowest.
Click to expand...

Bingo. ;)
 
Fools. Everyone knows that we'll be using 100% optical computers in ten years time.

At least it's clear that many of the participants in this thread have little to no experience with anything other than (Intel and AMD) x86 architecture-based systems ;)
 
I'm pretty sure you cannot make something exceed the speed of light.


Also, i read an article where some scientists in Colorado chilled some rubidium atoms to .0000000001 Kelvin and formed a Bose-Einstein condensate.....imagine having THAT cool your CPU:eek:
 
Wow. The beginning of this thread was bad enough... :rolleyes:

As devices shrink, the power density rises, even though the supply voltage drops. This is simple physics.

Though we may have something that is smaller and lower voltage, it's still drawing more and more current as the clocks rise and the internal resistances drop. Again, simple physics.

Thus we have something really tiny that puts out a shitpot of heat.

Processors are not going to get "cooler." If we continue along the trends that we have been since the inception of the integrated circuit, they will get hotter and we will need something more efficient at removing heat from them.

LN2 is not practical, nor is LH2, and it's explosive, and carries along with it "slight problems" for silicon.

Watercooling is nowhere near its apex for the normal home user. It is the most likely path of the cooling industry on a whole to follow as temps rise and sizes shrink.
 
Originally posted by Elledan
Fools. Everyone knows that we'll be using 100% optical computers in ten years time.

At least it's clear that many of the participants in this thread have little to no experience with anything other than (Intel and AMD) x86 architecture-based systems ;)
Click to expand...

lmfao 'Fools. 600 kb or RAM is all we'll ever need.'

I have seen some advances towards obtical components, but it will be a long time until this is a viable technology. It has a few... Limitations to overcome first.

Also, we've been using printed circuits for a few years now... I don't see THAT much research into killing off PCB.
 
I doubt liquid hydrogen and liquid nitrogen would be used for anything but fringe cooling. It is something you have to go buy or have delivered, and I know I wouldn't want to have to refill my computer every day or so. The equipment to produce liquid nitrogen and liquid hydrogen is so complex and so expensive that there is no way people would have on in their homes. I think air cooling is here to stay. You need some fluid that you can move over every component. You can do just that with fans. I can see liquid cooling going mainstream. And people would have to check their water level like they check the oil in their car.
 
Originally posted by ’m‚³‚ñ
lmfao 'Fools. 600 kb or RAM is all we'll ever need.'
Click to expand...
I'll keep my 640 Kb of RAM, thank you very much :p

I have seen some advances towards obtical components, but it will be a long time until this is a viable technology. It has a few... Limitations to overcome first.
Click to expand...
I don't know. Considering the latest couple of breakthroughs on manipulating photons, I get the feeling that we're pretty close already. Perhaps it won't be in 2014 that we'll be using optical CPUs, but electronics as it exists today will have to give way to spintronics, optical circuits or a combination hereof at some point.

Also, we've been using printed circuits for a few years now... I don't see THAT much research into killing off PCB.
Click to expand...
It's easy to mount stuff on a 2-dimensional, stiff piece of cardboard, plastic or other material. It's difficult to design 3-dimensional circuits and stack the components together in the right way.

It's pretty much the same reason why all of our cities are still pretty much 2D, with few attempts to stack things on top of each other, let alone interconnect them.
 
Originally posted by Elledan
I don't know. Considering the latest couple of breakthroughs on manipulating photons, I get the feeling that we're pretty close already. Perhaps it won't be in 2014 that we'll be using optical CPUs, but electronics as it exists today will have to give way to spintronics, optical circuits or a combination hereof at some point.
Click to expand...

Have you heard of the Hydrogen Powered car? I heard a story (I call bullshit on this story, but still) about a guy who made a working prototype of a hydrogen powered car. The Oil industry paid the guy LOTS of money for this thing, and promptly destroyed it so nobody could use it. While this story is most likely the remenants from a horse's ass, the idea behind the story could possibly be applied to certain technologies today.
 
You must log in or register to reply here.
Back
Top