Reports on AppleInsider and MacRumors say a liquid cooling system is possible for the PowerBook G5. Personally I'm a tad skeptical. Discussion?
P.S. Don't some Mac laptops already use liquid cooling?
P.S. I can't wait to see what iSegway thinks of this latest development!
Comments
Originally posted by ryaxnb
Reports on AppleInsider and MacRumors say a liquid cooling system is possible for the PowerBook G5. Personally I'm a tad skeptical. Discussion?
P.S. Don't some Mac laptops already use liquid cooling?
P.S. I can't wait to see what iSegway thinks of this latest development!
I am also skeptical. A die shrunk, lower voltage G5 would probably run effectively at 1.8GHz in a laptop without an ornate heat sink. Water cooling is neato and all, but it will cost money, and putting it in a laptop will cost even more money. Also, if your laptop was in freezing weather long enough for the water to freeze, you would have ruined your laptop (unless you have antifreeze in your laptop or you use some liquid that freezes at a much lower temperature).
i just dont see it.
they would going backwards actually.
I really don't see a problem with it. If it means G5's in laptops sooner, then good
-Dan
It either works, or doesn't. Apple will weigh the cost/benefits and implement it if it makes sense.
I think people imagine a tester sitting down to test try the new laptop and starts to shriek, "OH MY GOD, [sparks shooting out all over] SOMEBODY PUT SOME DAMN WATER IN THIS COMPUTER...OWWWW...IT HURTS...GET IT OFF ME!!!"
It won't be like that
1) Liquid cooling does not mean water cooling. As the AI article notes, one solution involves methane in both liquid and gaseous states.
2) Apple have used liquid cooling in their notebooks for years, and as AI notes again, so has IBM. It's always been part of a larger, multi-technology solution, but it's been there nonetheless. I've never heard of someone's Pismo suffering from a frozen cooling system.
3) The point of liquid cooling is to wick heat away from a difficult spot to a radiator, so that you can use a part that requires a heat sink in a place where there's no room for a heat sink.
4) Liquid cooling can (and in a notebook, should) be passive; which is to say that it's powered by the same heat it's supposed to wick away. This makes it quiet, if difficult to design.
The main complaints in the iSegway thread were about the initial assertion that liquid cooling was an unambiguous improvement in a tower, and that it was sensible as a sole means of cooling anything. The water-cooled Hitachi mentioned in both that thread and in the AI article is an bloated, heavy $3000 iBook with an onboard water tank, which is not a resounding endorsement of the technology.
However, liquid cooling certainly has its strengths, and one of its strengths is wicking heat out of confined and hard-to-reach places and diffusing it. In a thin slab of a notebook with two very heat-dense parts (the CPU and the companion chip) this might prove necessary. Certainly, the solution that made sense for the PowerMac G5 isn't going to work.
What if you could "way-station" waste heat into an aero-gel-insulated (or more practically, some space shuttle-ish ceramic tile) depleted-uranium core? Say store about 480 watt?hrs? In the course of a 12 hr service use, this DU "plug" would get obscenely hot (like 300-400 deg C?), but it is insulated to prevent any thermal damage to neighboring components (including your lap). How do you push that heat into the plug? You use some "volcano-ready" peltier cooler module (and perhaps a heatpipe) to channel the heat from "CPU & friends" to the DU core. (Hey, "core", sort of like the movie? ) Note that the peltier module isn't necessarily "cooling" the host, but pushing heat from the nominal CPU temperature to the intense temperature potential of the DU core.
So what happens after the 12 hr service period (or some time before then when the computer task has been completed)? Well the DU core has to be cooled somehow. Maybe you dock the laptop into something which flushes liquid through the core? ...or maybe the core is auto-mechanically swapped into the dock in exchange for a "fresh" core. The hot core is then cooled at leisure using the dock's heatsinking network. Mind you, this core could be something like a 2" DIA by .25" puck.
Sounds crazy! Sounds impractical! Well I did say it was an "absurdo-bizzarro solution".
I know the basic idea: It takes a lot of heat energy to lift matter to a more energetic state--liquid to gas, for ex. That is, it takes a lot more heat to lift water the 2° from 211° to 213° than from 209° to 211°. (again, someone can say this better or more correctly)
Amorph, are you sure the PowerBooks use liquid cooling? I've heard it was gas...which would still make it 'fluid' cooling I guess.
So the Powerbooks may use a substance that is a liquid when its cold but turns to a gas--taking away a lot of heat--as it passes by the processor. In other words, both you and Amorph could be right.
Also, liquid cooling isn't nearly as silly as you might think. Tom's Hardware did a decent how-to for installing water cooling on a desktop PC about a year ago.
http://www6.tomshardware.com/howto/20020701/index.html
Shrink things down, use tubing that won't corrode or break, and fill it with a high-tech liquid well suited for the task of heat absorption, and it just might work. This technology has already been used successfully in PC laptops.
that is exactly how the air-conditioning in your car works ... in physics, it's called a "vapor-cycle machine. ... The problem with that in a computer would be the compressor (required to return the gas to a liquid state)
Originally posted by reynard
Could someone who is better versed in physics explain the advantage of using a liquid that turns to a gas a relatively low temperature as a cooling medium?
I know the basic idea: It takes a lot of heat energy to lift matter to a more energetic state--liquid to gas, for ex. That is, it takes a lot more heat to lift water the 2° from 211°F to 213°F than from 209°F to 211°F. (again, someone can say this better or more correctly)
So the Powerbooks may use a substance that is a liquid when its cold but turns to a gas--taking away a lot of heat--as it passes by the processor. In other words, both you and Amorph could be right.
In general the energy temperature to go between any tow temperatures with an equal difference is equal.
Specifically ∆H = mC(T)∆T where ∆H is the change in energy, m is the mass, C(T) is the specific heat and ∆T is the final - initial temp. The specific heat is actually a function of temperature at constant volume or pressure but for small changes you can assume it constant.
The exception to this is when the liquid (or solid) actually goes through a phase change since you have to overcome certain intermolecular forces to do so. That's quantified by the heat of vapourisation and in many cases can be quite substantial.
The reason for using vapourisation in a cooling system is you keep the liquid/gas at a constant temp while removing a substantial amount of energy. This allows you to maximise your heat transfer driving forces, which is a function of the inlet and outlet fluid temps, and thereby your overall heat transfer which allows you to use smaller areas for the heat exchangers. That can be quite important particularly in a highly confined space.
That's a very simplified version of the theory. There are also a few disadvantages with that sort of method too and to really explain it I need diagrams but basically if you can understand there are 2 heat exchangers and you're trying to maximise your overall driving forces and heat transfer you can work out why having a plataeu region for temperature where energy is still removed might be advantageous.
Originally posted by reynard
Liquid cooling doesn't make the heatsink smaller; it just moves it.
Yes and no. Heat transfer coefficients for air heat exchangers are quite awful compared to liquid cooling systems. You can achieve the same result using a liquid cooling system that is smaller in size generally. Computers use a sealed system though which makes it not work quite as normal liquid cooling though since it is actually 2 HXs in series, a liquid cooled and an air cooled one. Obviously the air cooled one is going to limit the operations still.
Edit: Because I forgot American's usually use old units. Seriously switch to Celcius or Kelvin people
Originally posted by Randycat99
Randycat's absurdo-bizzarro solution:
What if you could "way-station" waste heat into an aero-gel-insulated (or more practically, some space shuttle-ish ceramic tile) depleted-uranium core?
...
So what happens after the 12 hr service period (or some time before then when the computer task has been completed)? ... Maybe you dock the laptop into something which flushes liquid through the core? ...or maybe the core is auto-mechanically swapped into the dock in exchange for a "fresh" core. ...
Sounds crazy! Sounds impractical! Well I did say it was an "absurdo-bizzarro solution".
That's crazy man. Crazy. An ablator would be a much better solution than depleted uranium.
The real solution is thermo-electric materials! With advances in nanotechnology making thermo-electric materials viable, the heat can simply be converted into electricity that recharges the battery. In fact, thermo-electric materials will just be embedded onto chips themselves in the future, this keeping them cool and returning the waste heat as electricity...
Originally posted by Telomar
Well actually the energy to go from 211° - 213° is basically equal to the energy to go from 209° - 211° since your specific heat will remain relatively constant.
I don't think that is quite right. While the specific heat may remain constant while traversing the boiling point (assuming the original reference was 212 deg F and we're talking about water), there is the heat of phase change to contend with. That is where a big chunk of energy absorption occurs.
Originally posted by Randycat99
I don't think that is quite right. While the specific heat may remain constant while traversing the boiling point (assuming the original reference was 212 deg F and we're talking about water), there is the heat of phase change to contend with. That is where a big chunk of energy absorption occurs.
I assumed it was talking in celcius so didn't even consider the phase change at 212°F. The answer covered heats of vapourisation anyway it just wasn't very neat.
See the importance of units people? A lesson to all future engineers.