CPUs you can Barbecue on?

[pave_spectre]

I recently read this (http://australianit.news.com.au/common/print/0,7208,6666477%5e15864%5e%5enbv%5e,00.html) article and saw this
Two G5 chips apparently produce a prodigious amount of heat. The G5 desktops have no fewer than nine cooling fans, but Apple claims they are quieter than the older G4 systems.
This thought may take a while but I felt it was worth sharing so please, bear with me.
Combining this with my recent attempts to get a pair of old computers up and running I started to think about comparative requirements in regards to cooling. One, a dual P90 system has no HS/F on either processor, while the second a P133 has a tiny HS/F that does not even have the fan wired up.
Comparing these to the needs of my main Athlon XP system, got me thinking about what future processors will require in terms of their cooling technology.
With the "soon-to-be" mass release of new 64 bit chips from Intel and AMD and the current Apple G5(as quoted above), probably the one thing I have seen no real information on, is just how much cooling they (and even future chips [128 256?]) will actually require.
Barring solid evidence to the contrary, it seems to me that without some significant leap in materials and electrical sciences we could end up needing to house future desktop PCs in large server towers just so we can contain all the cooling equipment needed not to turn a PC into a small expensive grill.
Will future processor boxed sets come with complete with peltier elements, or the manufacturer recommendations for their preferred water cooling systems? Will future overclockers be reduced to something like this (http://www.heatsink-guide.com/peltier-sandwich.htm) Peltier Sandwich?
If heating does become a significant factor where does the future lie for things like rackmount servers? Will home computer usage dwindle as electrical bills rise due to power consumption of your choice of cooling?

This whole thought might be considered as an adjunct to
this thread (http://www.pcguide.com/vb/showthread.php?s=&threadid=23693)

[deddard]

Hopefully the future lies in a different direction than the current capacitive/electron based processing. Magnet or even light based computers should significantly reduce the heat output. Newer ways to deal with electrical circuits (Hall effect perhaps?) need to be persued as well.

[Fruss Tray Ted]

In that Peltier sandwich, why does he want to use some type of insulation between the 2 heatsinks? Insulation is a method of retaining heat not dissipating it. It looks as if the edges of the cores would get rather hot due to the retention or reflection of heat.

I myself would be looking to some type of heat dissipator or spreader, even leaving the space open to receive airflow?!? Foam, epoxy, and the other forms he mentions seem inferior to something more fluid that would more readily give up it's accumulated heat to the adjacent heatsinks. Something like a gel that won't break down from high temps but thinner than hight temp silicone rubber or things of that nature.

Smell what I'm stepping in? :eek:

Making an O-ring or grommet to surround the perimeter of that arrangement... oohh and some piping... a mini radiator... 50/50 mix of antifreeze... :p oops, I mean :cool:

[Steve]

I don't know if they're hot enough to barbecue on, but those AMD chips are hot enough to fry an egg...:eek:

Sunny side up, please (http://www.handyscripts.co.uk/egg.asp)

EDIT: I know, it's an old one, but I thought it was funny...;)

[Fruss Tray Ted]

Steve,
Is that Maple Syrup? I'been known to boil abit meself, don't you know...

[Steve]

Maple syrup! Heck no. That's genuine HP Sauce (http://www.higherdesign.com/brownsauce/hpsauce.htm) !


But I bet maple syrup would be better...;)

[Vic 970]

a chap I saw on telly a while ago said that pretty soon they will be building houses around pc's. as the pc will be in control of the household functions, phone, telly, alarm etc.etc.

perhaps they will pipe it up and run the central heating off it !!!

[superdrumr]

eventually you will simply need to put the tower in a closet, right next to an air conditioning vent

[Budfred]

No, eventually computers will be either organically based and put out very little heat or they will use something like laser processing and produce almost no heat. Right now processors have gotten hotter because they are still using old tech, but if they are going to continue getting faster, they will have to get much more efficient and cooler...

Ironically, the first computers had to be air conditioned to keep the tubes and other electronics working. There are supercomputers that are cooled by subzero temps all the time. It is only recently that desktops have gone back in time by requiring all that extra cooling....

[pave_spectre]

I can just imagine an organic based computer growing like a fungus on the wall.
I cant see that organic or laser based processing or anything else similarly advanced will be around any time soon, so I guess we will just have to tolerate the heating effect. Nice in winter but danged inconvenient during summer.

[jeeza]

I am new to this forum and am posting herewith my first message.
I would like to submit an idea which I hope may prove to be fruitful.
As I understand, CPU's generally produce lots of heat, especially the Apple's CPU's you mentioned.
All this results in quite some thermal energy produced and wasted ; furthermore lots more energy are used just to cool these CPU's.
The overall result is lots and lots energy just warming up the environment without being actually put to any use, and we know about this environmental problem which is becoming more and more serious.
The solution I think would be to have some way of storing this energy instead of dissipating it.
Batteries exist which store electrical energy. Why not develop batteries which are able to do the same thing for thermal energy ?
Basically, what would be needed is a medium which is able to do this and a materiel which can effectively shield this medium from the environment.
Imagine a pc with, next to the barely lukewarm CPU, some kind of little box which you either can take out and "empty" at a proper time and place, or in some way, at a moment when the thermal energy can be used, connect to some output channel.
I think this kind of scenario would be ideal.
I believe thermal batteries are needed and would be useful besides electrical ones ; this is the consideration sustending this idea I am submitting.
I would appreciate if you could give some comments if you have any ?

[Budfred]

Welcome to http://www.pcguide.com/ubb/pcgubb.gif

The idea is basically sound and the reason I know that is that thermal batteries have been around a lot longer than electric batteries.... Shortly after the invention of fire, humans figured out that rocks got hot and stayed hot for a long time even after the fire was removed. They used this for cooking and heating purposes.

Solar homes today often use this principle to store thermal energy in a number of different ways and to feed it back at cool times. The amount of heat released by a single or even multiple CPUs is pretty negligible when compared to the amount of heat released by a car in a ten minute ride down the highway. The sun produces huge amounts of heat that are either absorbed by various objects or simply reflected back out into space.

The reason that engines and CPUs produce so much heat is that they are inefficient users of energy. We probably need to focus more on finding ways to produce less heat with more efficient engines and CPUs than we need to store and use that heat in other ways.

[pentachris]

A better way to remove heat is on the horizon. Imagine something like aluminum foil that you would put between your heatsink and chip instead of thermal paste or a transfer pad. Start a very fast reaction at one edge of the foil, and it quickly melts and then hardens, bonding the two together with heat conductivity 10 times better than that of Arctic Silver's Ceramique.

Here. (http://www.theinquirer.net/?article=10678)

[jeeza]

Thanks for the welcome.

I know heated stones have been and are been used, but this not my idea of thermal batteries, as you can easily imagine. :)
It goes without saying that we need something more adapted to our computerized and technologically developed environments.

I agree that, if you compare the heat generated by a CPU to that generated by a short car ride, it is negligible. Nevertheless, comparatively negligible as it is, it is still bothersome and disturbing to those who use computers (and I suppose to computers as well) and it must be disposed of, as proved by the fact that they are being cooled, by up to 9 fans in the new Apple machines as stated in the first post. This indicates to me that some better way *is* needed to dispose of this excess heat.
Logically, what would make more sense than a cheap and simple system of storing this heat ?

We have known for a long time that engines and now CPU's as well are using energy in an inefficient way. If it is possible to find means to improve this efficiency without too much delay then it should simply be done, otherwise it might be more easy, in the meanwhile, to use thermal storage devices, which shouldn't take any time worth mentioning to be developed.

[Budfred]

I am afraid that last statement is the problem. Heat is a very dynamic form of energy. Creating a device that would be small enough to use inside a computer (and especially inside a laptop) to store heat energy so that it can be used later is a very daunting task indeed. It would probably require creating a vacuum around the storage media and having a way to keep the movement of the heat in one direction until it is required to be used in the other. The only thing I have ever heard of that might make this feasible is a type of solar cell that converts heat into electricity. If that technology was used it could allow for heat to be stored in a more conventional battery which is a technology that is readily available. If feasible, it certainly could be a major boon for laptops by cooling them and recharging them at the same time. Of course, it wouldn't fully recharge, but it would probably extend battery life....

[superdrumr]

there is only one problem with all of the ideas presented. cost

[Budfred]

Actually, the heat absorbing to electricity silicon that I am talking about could be manufactured quite cheaply if someone developed the idea for use in a laptop or other system. It would be quite a bit more expensive for a car, but then it would produce a lot more electricity too and could be set up to gather solar heat as well.

[superdrumr]

The cost would be in initial R&D, plus any new technology is expensive, that is just how a capatilist economy works. However it may come down fast. Nevertheless with the tech industry in bad shape and all energy focused on speed, there is not enough attention paid to cooling

[jeeza]

Sorry for having delayed so long.
I think more S&D will go in this direction when the basic thinking in our society has changed into a more waste-conscious way of thinking.
As the late Buckminster Fuller pointed out in one of this books, half of all available sulphur is borne in the air and is produced partly from human industrial activity... sooner or later, it will become profitable to look for ways to extract our sulphur needed for industrial purposes from the air.
In the same way, at some moment it will become profitable to store and use as an energy source the thermal energy produced by our machines and household appliances.
As for your suggestion about heat conversion to electricity and storage in batteries, I suppose this would be the most obvious solution at hand and maybe the easiest to develop.

[jeeza]

Today I found this article, "Scientist measure coldest temperature ever", on http://rednova.com/news/stories/2/2003/09/12/story001.html .
They produced in their lab a temperature of half a billionth of a degree Kelvin.
So it is actually possible to create an unlikely temperature differential.
I wonder if the practical realization of a heat energy storing medium capable of keepîng a temperature of several hundred degrees, maybe more, inside a calorie-tight container, for a very long time, wouldn't be a possibility in a not so far future ? I can almost see it - something like a thousand degrees C inside a small container.
A few days ago I read somewhere (but for the life of me I couldn't rememvber where - I thought it was on http://www.kuro5hin.org but couldn't find it back) about a traditional system which is used in Korea, named "dondol" or something like that, used to heat part of their houses in a very efficient and economical way. The smoke coming from the stove used for cooking, and which is located lower than the floor of the room it is going to heat) is led in a cleverly intricate fashion through conduits under the floor to the other side of the house where it exits through the chimney.
There was said to exist a house which could hold the warmth produced by a fire burning for just two hours, for up to three months. This house is supposed to have been destroyed a long time ago, but scientists have tried to imitate it, and came up with a model which was able to keep the warmth for three days...