How does Apple do cooling, SO well?! :)

Hello, good evening, and welcome to my thread!.
-----------------------------------------------

Greetings people; I wanted to know something, and it could be a twist of rhetoric, but:

How do Apple manage to cram SO SO many exothermic devices inside their computers *AND* keep them well ventilated *AND* keep them SO SO quiet *AND* prevent them overheating (and keep them from looking the size of the Empire State Building)... when all those "think backward" guys over at the PC design labs/<insert inane name of overclocking brand hardware here> are pumping water/liquid refrigerant through oversized copper cooling heads, coupled with noisy and clunky, peripheral heat exchangers?!

I'm both astounded and impressed at the same time at how Apple manage it, but bemused by the very idea of pumping LIQUID through an electronic device, SO close to the motherboard.

What with Core 2 Duo/Extreme etc and 45nM dies emerging, it seems rather redundant and also excessive, to need/want to overclock anything, these days, so WHY do they (and why do they go to such lengths, but not only that - WHY do they show off all their mods, as if it were a teenager's car)

Colour me cynical, but I put it down to lack of taste, combined with a subtle blend of "cracking a nut with a sledgehammer". Has technology stood still in the land of all things DIY-PC, or am I imagining this?.

I tried to get my head around it, but it still astounds me - maybe someone can explain the logic of neon lights and pond pumps, in a COMPUTER.

Thanks in advance.

glosswhite said:

Colour me cynical

Click to expand...

OK.

It's actually quite simple. They put neon lights in because to make up for the software they have to make the hardware look good. The software primarily meaning Windows.

smurfy said:

OK.

It's actually quite simple. They put neon lights in because to make up for the software they have to make the hardware look good. The software primarily meaning Windows.

Click to expand...

Granted on that, but there must be some phallic symbolism/inadequacy, buried not TOO deep within this mess of pipes and garish UV paintwork.

So, basically, you are saying that it is the computing equivalent of getting a 1983 Ford with a 1.0 Litre engine, scraping all the rust from its wheel arches, and plastering it with shiny fibreglass skirts, and underglows?.

Thought as much LOL!! (excuse me 'LOL'-ing; having a break from sensible phrases, as I am tired)

If Apple were rubbish, then it would make NOT wanting to be a fanboy, much easier.

glosswhite said:

I'm both astounded and impressed at the same time at how Apple manage it, but bemused by the very idea of pumping LIQUID through an electronic device, SO close to the motherboard..

Click to expand...

Apple has used liquid and a small 5" x 5" radiator to to cool the G5 PowerMac. I have seen the cooling unit, and the repair person told me it holds about a cup of liquid. There are also eight or nine fans cooling the unit.

I don't know how Apple is cooling the other units. I will know what is in the Mac Pro, when I get it.

rman said:

Apple has used liquid and a small 5" x 5" radiator to to cool the G5 PowerMac. I have seen the cooling unit, and the repair person told me it holds about a cup of liquid. There are also eight or nine fans cooling the unit.

I don't know how Apple is cooling the other units. I will know what is in the Mac Pro, when I get it.

Click to expand...

The G5s needed SO MUCH cooling! That's part of the reason they switched to Intel! It was getting to be humorous!

Yeah, the switch to Intel was mostly about efficiency (though pure speed was probably also important). There never was (and never will be, now) a G5 laptop because the cooling apparatus would have been bigger than the case it had to fit in!

I had a simple, but possibly revolutionary idea last night, that could possibly cool the latest cores, to previously un-thought of temperatures. It doesn't involve heatsinks or cooling devices, either .

The *principle* is simple, but I don't know if it would work. Hmmm. I am no processor designer, but logic dictates that my theory should also work in practice... "should" .

I am unsure about revealing my idea - could bag me some £££ - any ideas?.

glosswhite said:

I had a simple, but possibly revolutionary idea last night, that could possibly cool the latest cores, to previously un-thought of temperatures. It doesn't involve heatsinks or cooling devices, either .

The *principle* is simple, but I don't know if it would work. Hmmm. I am no processor designer, but logic dictates that my theory should also work in practice... "should" .

I am unsure about revealing my idea - could bag me some £££ - any ideas?.

Click to expand...

No problem - would love to contribute lots of money. Hold your breath for just a few minutes while I mail it to you.............

MAcbooks and MBP's are kept cool by using your lap as a heatsink... ok that was a joke.

Apple TV is kept "cool" by using the entire lid as a heatsink, and that is NOT a joke.

Generally, you'll find Apple's machines clocked just a teeny weeny bit slower that the leading/bleeding edge Dell's for example, which helps keep the temps down.

However most of it is down to design. When you buy a Mac, you're paying for Apple Design.

Microsoft could learn from this :/

Yeah, I have a imac G5 (not intel version)... This is my biggest problem, the fan noise. The fan is what seems to be constantly running at the highest, or near highest speed. I can't seem to fix it..

Sandwichman said:

Yeah, the switch to Intel was mostly about efficiency (though pure speed was probably also important). There never was (and never will be, now) a G5 laptop because the cooling apparatus would have been bigger than the case it had to fit in!

Click to expand...

Actually cooling was not the primnary factor, it was power consumption and although the two are related, the design of the G5 was never intended for portable use.

Steve's famous words were 'power per watt', which is where the Core Dup design really shines.

glosswhite said:

I had a simple, but possibly revolutionary idea last night, that could possibly cool the latest cores, to previously un-thought of temperatures. It doesn't involve heatsinks or cooling devices, either .

The *principle* is simple, but I don't know if it would work. Hmmm. I am no processor designer, but logic dictates that my theory should also work in practice... "should" .

I am unsure about revealing my idea - could bag me some £££ - any ideas?.

Click to expand...

Does it involve putting the Mac in a freezer?

If so, I tried it, it doesn't work, sorry.

smurfy said:

Does it involve putting the Mac in a freezer?

If so, I tried it, it doesn't work, sorry.

Click to expand...

Do you ever post anything of any value here?

smurfy, that was indeed, a pointless and stupid comment. No, it has nothing to do with freezers, and I am pretty sure noone has anything like this idea.

Well here it is:

The dies on chips are shrinking ok - why don't they just INCREASE the die *area*, whilst keeping the 65nM/45nM track width, and space the semiconductor tracks out, with a wider spacing, so that the heat is dissipated over a wider area?. Surely if the ACTUAL DIE was doubled or tripled in size, then the surface area would be cooler, due to there not having all the tracks condensed down in a tighter packed space.

Just my idea - anyone know if that would work?. A bit like having a fireplace FULL of red hot coals, then taking them out and splitting the coals into 1mm square blocks, and spreading them over 100 Metre square area... surely the heat dissipation would be better?!.

Cooling Idea

glosswhite said:

The dies on chips are shrinking ok - why don't they just INCREASE the die *area*, whilst keeping the 65nM/45nM track width, and space the semiconductor tracks out, with a wider spacing, so that the heat is dissipated over a wider area?. Surely if the ACTUAL DIE was doubled or tripled in size, then the surface area would be cooler, due to there not having all the tracks condensed down in a tighter packed space.

Just my idea - anyone know if that would work?. A bit like having a fireplace FULL of red hot coals, then taking them out and splitting the coals into 1mm square blocks, and spreading them over 100 Metre square area... surely the heat dissipation would be better?!.

Click to expand...

I think that idea would help with cooling, but as a side effect processing speed would drop dramatically, because path length would be so much longer.

glosswhite said:

smurfy, that was indeed, a pointless and stupid comment. No, it has nothing to do with freezers, and I am pretty sure noone has anything like this idea.

Well here it is:

The dies on chips are shrinking ok - why don't they just INCREASE the die *area*, whilst keeping the 65nM/45nM track width, and space the semiconductor tracks out, with a wider spacing, so that the heat is dissipated over a wider area?. Surely if the ACTUAL DIE was doubled or tripled in size, then the surface area would be cooler, due to there not having all the tracks condensed down in a tighter packed space.

Just my idea - anyone know if that would work?. A bit like having a fireplace FULL of red hot coals, then taking them out and splitting the coals into 1mm square blocks, and spreading them over 100 Metre square area... surely the heat dissipation would be better?!.

Click to expand...

I think the boffins at AMD and Intel etc have thought of this.

The problem is that larger circuits require more voltage, which consumes more power and creates more heat. So there is a payoff for having smaller wafers and more transistors in an ever shrinking space.

The real issue is getting the heat out of that tiny space and onto the heatsink effectively, but unfortunately your idea would not work, because something like a 65nm CoreDuo increasing to say 130nm would need almost twice as much power to run at the same ghz.

glosswhite said:

smurfy, that was indeed, a pointless and stupid comment. No, it has nothing to do with freezers, and I am pretty sure noone has anything like this idea.

Well here it is:

The dies on chips are shrinking ok - why don't they just INCREASE the die *area*, whilst keeping the 65nM/45nM track width, and space the semiconductor tracks out, with a wider spacing, so that the heat is dissipated over a wider area?. Surely if the ACTUAL DIE was doubled or tripled in size, then the surface area would be cooler, due to there not having all the tracks condensed down in a tighter packed space.

Just my idea - anyone know if that would work?. A bit like having a fireplace FULL of red hot coals, then taking them out and splitting the coals into 1mm square blocks, and spreading them over 100 Metre square area... surely the heat dissipation would be better?!.

Click to expand...

One of the advantages of reducing the die size is reducing the track length, thus decreasing the distance it takes for electrons to travel from point A to point B. It could be done, at the sacrifice of performance.

Of course there are so many ways to cool CPUs, that there really is no point in doing this. Clearly air cooling is still effective and can be implemented in so many different form-factors.

The guys that are over-clocking, water-cooling and customizing machines are doing so for sport. It's to see how far you can push the envelope. It's akin to people who build hot rods. Sure, they don't have the beautiful, simplistic design of the Mac, but that isn't their design goal either. To each his own, it might not float your boat, so you buy a Mac. Many people (myself included) wouldn't touch Apple hardware if I could legally run OS X on something else. My personal belief is that form should always follow functionality. Apple loves to remove useful keys/ports/features to streamline design, which is something I can't wrap my brain around.

I have seen a few non-intel G5s get extremely loud and some power supplies die eventually after only a year or so. But in general, macs seem to be better than PCs in terms of cooling and quietness

cwa107 said:

One of the advantages of reducing the die size is reducing the track length, thus decreasing the distance it takes for electrons to travel from point A to point B. It could be done, at the sacrifice of performance.

Of course there are so many ways to cool CPUs, that there really is no point in doing this. Clearly air cooling is still effective and can be implemented in so many different form-factors.

The guys that are over-clocking, water-cooling and customizing machines are doing so for sport. It's to see how far you can push the envelope. It's akin to people who build hot rods. Sure, they don't have the beautiful, simplistic design of the Mac, but that isn't their design goal either. To each his own, it might not float your boat, so you buy a Mac. Many people (myself included) wouldn't touch Apple hardware if I could legally run OS X on something else. My personal belief is that form should always follow functionality. Apple loves to remove useful keys/ports/features to streamline design, which is something I can't wrap my brain around.

Click to expand...

Half the joy of using a Mac, is using *A MAC*!!

Shame on you