Why not Water cooling? - Page 6  

You think this is nitpicking? That's all you have left? In your own mind you have proven something, perhaps. Why don't you let others here decide what has been proven in this topic?



That was debunked quite easily, despite your efforts to not acknowledge it. "Efficiency" has a nasty habit of getting lost when you have multiple elements in a system that work in a series. That said, the efficiency of the system cannot exceed that of the weakest stage in the chain. Since a water-cooled system typically relies on an air-cooled stage at the very end of the chain, the air-cooled part will determine the baseline efficiency. Taking the system as a whole, such a system must inherently be less efficient than a simple air-cooled system (unless you wish to argue that your water-cooled components have "negative efficiency"). There's more stuff in a chain. "More stuff" should tip you off that efficiency will likely be less than something else with "less stuff". The simpler something is, the more efficient it tends to be. That's a pretty reliable guideline.



How can I be the one "crying" when you are the only one here being so emotional?


Ahh yes, the secret condition appears. You envision somebody using a "copper rod" to duplicate the function of a heat pipe? Tell me, why would somebody use copper in such a way? It's a useless comparison because nobody in there right mind would use copper in such a manner in the first place! Why not utilize a more sane comparison with a heat pipe vs. a copper plate that forms a direct junction between CPU and aluminum heat sink? That should lead you to the conclusion that the whole reason you use a heatpipe is not because it is "better", but because you are stuck in a situation where you cannot physically put that heatsink (with integral copper plate) directly on the CPU. It's an application issue, not a performance issue.


I imagine you can only find humor in all of it because you are unable to functionally understand it in the first place.

To put this whole premise into an analogy:

A car is "better" to take a 5 mi trip to the supermarket, but does that make it "more efficient" than a bicycle? Efficiency and logistics are 2 different things. You can make compelling arguments to use the bike or car- it all depends on where you want to go and what you want to do.

What I was saying is where are you going to put the radiator. However... how about the basement or closet? It would be a bit better there.

A meaningless example. A heatpipe is designed to work with water as a vital component. Most things don't work when you remove a vital component. However, if you truly understood why a heatpipe works, you would realize that it technically could still work with air as the medium (albeit, under extenuating and impractical circumstances, but no more nonsensical than suggesting the use of a copper rod as a heat conduit). If it were possible to contain air under great pressure and low temperature in a liquid phase within the heatpipe, it would cool the hell out of whatever is connected to it far better than water.

My point was that Apple would have more freedom in the design of the case if they were able to have a cooling system that can be moved around without disturbing the design of the motherboard and internal components. The radiator can be placed in the back of the case, the top, side, wherever Apple wants it to be. The basement or closet is probably not a good idea but if you believe it, you should suggest it to Apple.

After sitting and reading through all the pages of this debate, it really seems that iSegway is trying to stir up a heated argument. I have an engineering background myself and many of the replies to his posts have been very detailed and factual. Liquid-cooled systems ARE better at removing heat from a specific area quickly. But the truth is unavoidable, that heat must be dissipated into the air. Since you move the heat away faster, you must dissipate it faster to keep the cooling cycle functional. The only way to dissipate lots of heat quickly are:

Large radiator, slower airflow
or
Small radiator, faster airflow.

Your car works this way. So does your home A/C system. Your refrigerator too.

All require pumps and fans. Granted, some of the articles posted detail systems that do not use pumps or fans, but these are also still experimental systems. For all we know, Apple has already tested them and are prototypes for G6 systems in a few years. Some of the systems out there that use freon as the cooling liquid have compressors to generate noise and take up space too.

The Apple design works because the system can cool the areas that run hot on an "as needed" basis. Remember, we're talking about CPU(s), power supply, RAM, north/south bridge, AGP card, PCI cards, and drive bays. Each of these areas will vary in temp as the machine performs different tasks.

Logically, liquid cooling in the current G5 system would not give you enough space back to really justify the cost. You still need conventional cooling for the components listed above. So in the area reclaimed by removing the massive heatsinks from the processors, you must install: heat-transfer block on the CPU(s), coolant lines, pump, and radiator. Judging from all the kits I've seen (and some I've actually played with) you gain no space. So the only real advantage to liquid-cooling would be overclocking...which is not something Apple is going to do. Maybe as power of the CPU increases and more radical cooling is required Apple will need to look into this, but for now it just isn't needed.

iSegway...you make some good points. But you also have, rather sarcastically, declared that others were wrong when they posted information contrary to your own opinion. And all your support for your claims has been from internet articles you have read (graciously posting links to, also.)

Some of the posts you have declared are wrong come from basic engineering principles and physics, posted by people who work with them on a daily basis, and to call these people wrong is just rude. I have worked on cooling systems for high-voltage television transmitters so I know all about many types of cooling systems. I also was ASE Certified in automobile repair (around 12 years ago) with specialty in air-conditioning and cooling systems. I currently teach broadcast engineering and computer-tech classes to college students.

Just don't be so quick to discredit someone else when they try to explain things to you, and when you get corrected on a particular item, learn from it and grow. Sounds like you got a good mind, just a bit closed.

Welcome to the thread. You'll fit right in, having lead your argument with a personal attack and self promotion.

Fluid cooling systems are being actively pursued by academics and are currently used in thousands of consumer and industrial products.

The ongoing research is fascinating and worthy of discussion. It is too bad that this debate is now more about the participants.
(And yes, I just contributed to that trend.)

Your question was irrelevant. I did give you an answer, however, just not one you would accept...but what else is new?


This sounds like an "open mind"?


He said a 400x factor was plausible. He didn't say that any ole heatpipe in any ole installation would achieve that.


A piece of copper shaped like a heatpipe- would that not be a "copper rod"? To be more precise, a slender, circular solid piece of copper. Is that not what you had in mind to compare to a heatpipe? If not, then what?


You would compare a heatsink and a heatpipe? Have you not the slightest clue that they have functionally different purposes? You are also twisting the meaning of my comments. What I did say is that the functional equivalent of a heat pipe (to be the interface between 2 mechanical regimes) would be a flat copper plate between a CPU and a conventional heatsink. The heatpipe is long and slender while the copper plate is flat. However, they both transfer heat from one location to another. In that sense, a heatpipe of appreciable length would not necessarily be "400x more conductive" than a thin, flat copper plate. A thin, flat copper plate would actually be hard to beat.


That is more your defensive reaction than anything I said. I only said that water-cooling inherently involves air-cooling at the end of the chain. Therefore, the overall properties, behaviors, and efficiencies are shared. They are not independent, and that makes it difficult to flatly say that one is better than the other (to your dismay), since in the water-cooling case they are both involved together.

Troll.

This thread is newly in danger of being closed. Stick to the subject and lay off the personal attacks.

You guys are so STUPID! I guess that's why you have Macs in the first place.

Stupid: Oblivious to the outside world.

Let's see... Koolance, Swiftech, Polarflo, Danger Den, Innovatek, etc. have retail water-cooling products available. Koolance has a case that Joe Blow can buy that is already set up for water-cooling, assuming you're not using a Mac.

If Joe Sixpack can change a HSF then he can water-cool. Yes, it's that easy now. I've been water-cooled for years!

My next move is a water-cooled peltier system. Which, if I wanted to, I could but off the online shelf! Polarflo doesn't have a pelt setup yet, so I'm going to toil in the dungeon until it's finished.

I really am a nice guy and don't "hate" Mac-lovers. My best man at my wedding still owns a Mac. I'm glad to see you guys looking out the windows and seeing the light about water-cooling.

Maybe you should be, um... "unregistered" for bumping this ancient thread?

Bad Noob Troll! No donut!

good god. this thread has been dead for half a year. how in the world do you crack monkeys even find this crap?