The conversation on this page has been archived and is no longer active.

24" iMacs run hot!

Message #1 - Posted 2007/06/07 - Mike Rosenberg

Jolly Roger wrote:

Honest mistake due to multitasking (it's been a busy day for me). Sorry for the confusion, really.

I think it's the doing of The Conrad Gang.

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://comedancing.biz> Ballroom dance shirts & gifts

Message #2 - Posted 2007/06/07 - Paul Nevai

Jolly Roger wrote:

You hard-headedly refuse to admit that the temperature of the COMPONENTS is what really matters, not the temperature of exiting air!

There is another issue to consider. This has nothing to do with the original subject. Namely, the temperature of exiting air is directly related to the temperature of the room where the computer is located.

I have 2 macs in my rather large study that are on 24/7 and the room is always warmer than it should be. In the winter I close all the heating/AC floor registers and in the summer I open them + I have a ceiling fan. Of course, the hard disks contribute to the heat as well.

/PaulN

Message #3 - Posted 2007/06/07 - Joseph Crowe

Paul Nevai wrote:

Jolly Roger wrote:

You hard-headedly refuse to admit that the temperature of the COMPONENTS is what really matters, not the temperature of exiting air!

There is another issue to consider. This has nothing to do with the original subject. Namely, the temperature of exiting air is directly related to the temperature of the room where the computer is located.

I have 2 macs in my rather large study that are on 24/7 and the room is always warmer than it should be. In the winter I close all the heating/AC floor registers and in the summer I open them + I have a ceiling fan. Of course, the hard disks contribute to the heat as well.

Actually, on my old G5 iMac, when things are really cranking with lots of disk accesses, the hard drive contributes more to the heat equation than the CPU, at least according to iStat Pro. It also contributes more to the fan noise. That said, my older Athlon based homebuilt runs hotter and noisier under both Windows XP and Linux. Technology advances and more powerful boxes have heat issues, but compute power per watt is still increasing.

/PaulN

Message #4 - Posted 2007/06/07 - The New Guy

No. I'm saying when a cooling design is good, it keeps the component cool.

Right, and it does that by transferring the heat to something else. In the case of the iMac, that something else is the air that gets expelled from the case.

Now you're starting to understand the problem. The point is for the CPU to transfer that heat from itself to the heat sink. The "something else" you just mentioned. If the heat sink is small (cheap) it gets hot fast, requiring a lot of air to keep it from overheating. But a lot of air means a fast moving fan meaning noise. So a good design would reject that. But if the heat sink is a good design, with adequate radiating area, that same CPU heat is transferred to a large area that even a small amount of moving air cools easily. And the heat sink does not get hot. If some of you would work with heat sinks this would be easy to understand.

Message #5 - Posted 2007/06/07 - The New Guy

The only documented heat failures I'm aware of involved solder joints in the Mac Plus, and any competent tech could fix that in 10 minutes.

I guess you don't get out much.
http://www.dssw.co.uk/sleepcentre/threads/new_imac_g5_heat_problems_sol ved.html

http://www.engadget.com/2005/03/04/imac-g5-overheating-problems/2

http://www.ehmac.ca/mac-ipod-help-troubleshooting/28275-overheating-pro blems-20-inch-imac-g5.html

I could go on and on. There are lots of people with heat issues. But anybody with the slightest knowledge of electronics knows that heat is not your friend and if you can drastically lower the temp, its a very good goal.

Message #6 - Posted 2007/06/07 - William Mitchell

The New Guy <replytogroup@here.thanks> writes:

But if the heat sink is a good design, with adequate radiating area, that same CPU heat is transferred to a large area that even a small amount of moving air cools easily. And the heat sink does not get hot. If some of you would work with heat sinks this would be easy to understand.

And of course, if there is a small amount of air carrying off all of the heat, then the air gets hotter. Therefore hot air coming out of a computer is evidence of good heat sinks.

That is the point you're making, isn't it?

Bill Mitchell
Dept of Mathematics, The University of Florida
PO Box 118105, Gainesville, FL 32611--8105 mitchell@math.ufl.edu (352) 392-0281 x284

Message #7 - Posted 2007/06/07 - The New Guy

Again, direct temperature readings confirm that the actual components are well within operating temperatures.

I notice you don't comment on the temperatures I posted. They show a well cooled hard drive running at about half the Mac temperature. 25C instead of 49C. Does that seem like good cooling to you? Or are you now going to say that 49 is just fine for a hard drive to operate at?

Fine - you want to talk numbers, here we go (yet again): WARNING: This may very well be the last time, because my patience with you is at an all-time low.

Duly noted. :)

The guy you quoted said: "My temps as i write this (idle) are 39C HDD, 31C CPU, 29C case in a 25C room."

That was before he improved the cooling. The resultant hard drive temp was 25C. I just tried it with a 250gb 7200 rpm IED drive. After putting a fan in front of it, it barely felt warm. I could hardly even tell it was operating. It was very pleasantly surprising. Fans rock. If your G5 tower had the air moving the opposite direction on that top compartment, with not so many air blockages with the framework of the case interfering with airflow, I bet it would be the same.

So his hard drive is running 14˚ C
warmer than the ambient room temperature. Sure, 39˚ C is cool, and nobody is arguing with that.

25C is cool. Which was very close to the ambient temperature. Which is my point all along. A properly cooled hard drive (simply a fan blowing on it with ambient temperature air, not hot interior case air) will be very close to ambient temperature.

But the machine was in a 25˚C room. The *difference* in ambient and component temperature is what determines the efficiency of a cooling system.

Yes - as I have shown on a well cooled system they will be very close, never 14C apart.

The temperature in my office right this very minute is 78.8˚ F (26˚ C), and the direct hard drive SMART temperature reading is 40˚ C. So my hard drive is running 14˚ C warmer than the ambient room temperature.

Well you can do better! :)

Note 1: If you'll recall when I did my temperature readings, my ambient room temperature was 77.6˚ F (25.3˚ C), and the hard drive exhaust air temperature was 88.7˚ F (31.5˚ C). So the difference between the ambient room temperature and that of the hard drive exhaust air was 6.2˚ C, likely because the fans were running faster than they are right now. This proves that there is no direct correlation between the temperature of the exhaust air and actual component temperatures without taking air flow speed and other factors into account.

Note 2: According to the hard drive manufacturer, Seagate, even 49˚ C is well within the stated operating range of 0 to 60˚ C. See for yourself:

They're just appeasing the manufacturers so they can continue with sloppy design. Anyway, is it good for business for Seagate to have drives working for 10 years? Sometimes, the ideal business model is not for a component to have a very extended lifespan. Not that anybody is going to use a drive that old, but everybody knows how awful it is for a drive to die (right in the middle of something important no doubt). This can now be largely avoided.

<http://www.seagate.com/docs/pdf/datasheet/disc/ds_barracuda_7200_10.pdf>

And regardless, as long as the temperature doesn't climb too close to the limit of the manufacturer's stated maximum acceptable temperature, it matters little. If it's within spec, the drive is fine.

Others would digress. Its just so easy to get the drive temps to near ambient room levels.

Message #8 - Posted 2007/06/07 - The New Guy

Previously, Jolly Roger wrote:

On 2007-06-06 14:45:05 -0500, The New Guy said:

Then they would have cool temperatures. Compare the temps of Jolly Rogers HD's to the SPCR example. 25 vs 49 C. They are about twice as hot!!!

The ambient room temperature in my office was about twice as hot as the SPCR guy's room temperature. I guess you figure that means nothing at all, huh?

How's that for brilliant Apple engineering?

The SPCR guy's hard drive was 14 degrees warmer than the ambient air in his room.

No - you didn't read the whole thing. That was before the cooling improvement. The it was near ambient levels.

Message #9 - Posted 2007/06/07 - The New Guy

Previously, Dave Balderstone wrote:

Previously, Ymir wrote:

Previously, The New Guy wrote:

Well that is true. But the fact is that hot components heat up small spaces rather well. And computers are small spaces. But you'd know this if you worked with hardware at all. Hot exiting air = badly cooled components. Period.

Umm. No. hot air *not* exiting == badly cooled components.

Here's an anecdote...

A few months ago, the air conditioning in our server room failed over the weekend. Our XServes shut themselves down due to thermal conditions when the ambient air in the room (we have a monitor that records it 24/7) hit 64C. (And there's a very cool - pun intened - error screen for when that happens)

At that point, everything in the room was at 64C. Racks, CPUs, walls, everything.

We opened the room as best we could (door and ceiling), and had high volume fans running for the week it took to get a replacement AC unit shipped to us here in the middle of the Canadian Prairies.

We had temperature sensors recording the air 6" behind the XServes. We kept that temp below 42C. If it had gone above 42C, we would have shut down everything and let it cool off.

No servers died.

We did end up, after the fact, replacing 7 of 28 CPUs, 7 of 56 drives (4 had their heads welded to the platter), 5 of 20 power supplies, 8 of 56 memory sticks, and 4 of 12 RAID controllers.

But we lost NO data, and were not late publishing the newspaper.

Apple makes some robust equipment.

Interesting story. Unfortunately, Apple doesn't make hard drives. So congratulate the hard drive manufacturers.

Message #10 - Posted 2007/06/07 - G.T.

"The New Guy" <replytogroup@here.thanks> wrote in message news:replytogroup-21C864.18155507062007@news.lga.highwinds-media.com...

No. I'm saying when a cooling design is good, it keeps the component cool.

Right, and it does that by transferring the heat to something else. In the case of the iMac, that something else is the air that gets expelled from the case.

Now you're starting to understand the problem. The point is for the CPU to transfer that heat from itself to the heat sink. The "something else" you just mentioned. If the heat sink is small (cheap) it gets hot fast, requiring a lot of air to keep it from overheating. But a lot of air means a fast moving fan meaning noise. So a good design would reject that. But if the heat sink is a good design, with adequate radiating area, that same CPU heat is transferred to a large area that even a small amount of moving air cools easily. And the heat sink does not get hot. If some of you would work with heat sinks this would be easy to understand.

God you're a fucking idiot. No one's arguing those last 3 sentences. You're making huge, absurd leaps of logic tying those last 3 sentences with the temp of the air coming out of a small exhaust vent. Your refusal to realize this is tiresome and trollsome.

Greg

Ticketbastard tax tracker:
http://ticketmastersucks.org/tracker.html

Dethink to survive - Mclusky

Message #11 - Posted 2007/06/07 - The New Guy

You hard-headedly refuse to admit that the temperature of the COMPONENTS is what really matters, not the temperature of exiting air!

One day it will dawn of some of you that if the temperature of a component rises and is not cooled with a heat sink, the temperature of the exiting air will be hot. The trick to this problem is to cool something with external air and by radiating the heat away from it as soon as possible. Or in the case of the hard drive, its the hard drive casing itself since we can't get air in there to cool the components that are heating the hard drive enclosure. So that's why a fan, using exterior ambient temperature air (not hot internal case air) works so admirably. A fan blowing on to it, not sucking hot air out of the case that just happens to wander past the hard drives.

There is another issue to consider. This has nothing to do with the original subject. Namely, the temperature of exiting air is directly related to the temperature of the room where the computer is located.

Yes - the 2 temperatures should be very close if the components are cooled well. As soon as something is getting hot, its a sign the heat sink is not up to the task. The same principle allows some heat sinks to cool CPU's with no fans at all. Just the slightest airflow is enough to whisk that heat away that is being radiated by all those cooling fins of the heat sink. And it follows that if one put a one of those large heat sinks, the air off that heat sink wouldn't be warm. And of course you'll notice every good heat sink these days uses heat pipes; almost all of them being expensive copper. So you can see why a manufacturer tries to shy away from expensive parts like that.

I have 2 macs in my rather large study that are on 24/7 and the room is always warmer than it should be. In the winter I close all the heating/AC floor registers and in the summer I open them + I have a ceiling fan. Of course, the hard disks contribute to the heat as well.

What model of Macs? It should be possible to replace the heat sinks in some models. Or by looking at the weakness of the design, one may rectify it.

For instance the G4 tower (specifically a dual 500 but others might be similar) can be cured so easily.
Disconnect the useless 4" fan.
Mount a 80 to 120 mm fan running at 5 volts (red and black wires off your Molex multicolored white plastic plugs that power things like DVD burners and IDE hard drives) so its silent, on the CPU heat sink. That will cool the CPU heat sink well. Now take apart the power supply and turn the fan so it blows in. Change the wires so its running at 5 volts (red and black). Now it will be almost inaudible. And it will be cooled much better now that you're not using hot interior case air to cool the power supply heat sink(s). The only downside to this scenario is you have to have the case open a crack to let the heat escape. But as long as you don't have a lot of dust, you're fine.

I saw a G4 tower 1.25 ghz the other day and the weak part of this design was mounting the hard drives so the air passes by them before being conducted to the CPU heat sink. Great for the hard drives, bad for the CPU. It has provisions for 4 hard drives (maybe more) so just don't use the ones at the bottom.

Fans have gotten much quieter these days so sometimes just replacing an older fan with a newer one can reap huge benefits. Sleeve bearings are the worst, single ball bearings are better, dual ball bearings are best.

Message #12 - Posted 2007/06/07 - The New Guy

But if the heat sink is a good design, with adequate radiating area, that same CPU heat is transferred to a large area that even a small amount of moving air cools easily. And the heat sink does not get hot. If some of you would work with heat sinks this would be easy to understand.

And of course, if there is a small amount of air carrying off all of the heat, then the air gets hotter.

The air temperature will rise but if the heat sink has a large radiating area (fins) that rise will be minimal. The better the heat sink, the less the temperature rise. Since almost any fan gets noisy above 7 volts, that means you're not dealing with a lot of airflow to arrive at a cooling goal. Which means you need a heat sink with lots of radiation area.

Also I don't think "small amount of air" is really appropriate here. A 120mm fan, even running at 7 volts moves a significant amount of air. And that air is very effective at cooling if positioned in front of something that gets all of it instead of just spread around like in a typical PC case. Ducting can work wonders.

Therefore hot air coming out of a computer is evidence of good heat sinks.

Actually the opposite. Go to a computer store and look at their best cooled computer - after talking about it with a salesman for a moment. You will see. Remember - the best cooled computer - not the quietest or the best performing. Both of those make it more of a challenge to cool a computer.

That is the point you're making, isn't it?

I think I just explained it again.

Message #13 - Posted 2007/06/07 - The New Guy

Previously, The New Guy wrote:

Previously, Dave Balderstone wrote:

Previously, Ymir wrote:

Previously, The New Guy wrote:

Well that is true. But the fact is that hot components heat up small spaces rather well. And computers are small spaces. But you'd know this if you worked with hardware at all. Hot exiting air = badly cooled components. Period.

Umm. No. hot air *not* exiting == badly cooled components.

Here's an anecdote...

A few months ago, the air conditioning in our server room failed over the weekend. Our XServes shut themselves down due to thermal conditions when the ambient air in the room (we have a monitor that records it 24/7) hit 64C. (And there's a very cool - pun intened - error screen for when that happens)

At that point, everything in the room was at 64C. Racks, CPUs, walls, everything.

We opened the room as best we could (door and ceiling), and had high volume fans running for the week it took to get a replacement AC unit shipped to us here in the middle of the Canadian Prairies.

We had temperature sensors recording the air 6" behind the XServes. We kept that temp below 42C. If it had gone above 42C, we would have shut down everything and let it cool off.

No servers died.

We did end up, after the fact, replacing 7 of 28 CPUs, 7 of 56 drives (4 had their heads welded to the platter), 5 of 20 power supplies, 8 of 56 memory sticks, and 4 of 12 RAID controllers.

But we lost NO data, and were not late publishing the newspaper.

Apple makes some robust equipment.

Interesting story. Unfortunately, Apple doesn't make hard drives. So congratulate the hard drive manufacturers.

Careless of me......didn't see that 7 drives were replaced (I'm guessing those were hard drives).

Message #14 - Posted 2007/06/07 - The New Guy

Previously, G.T. wrote:

"The New Guy" <replytogroup@here.thanks> wrote in message news:replytogroup-21C864.18155507062007@news.lga.highwinds-media.com...

No. I'm saying when a cooling design is good, it keeps the component cool.

Right, and it does that by transferring the heat to something else. In the case of the iMac, that something else is the air that gets expelled from the case.

Now you're starting to understand the problem. The point is for the CPU to transfer that heat from itself to the heat sink. The "something else" you just mentioned. If the heat sink is small (cheap) it gets hot fast, requiring a lot of air to keep it from overheating. But a lot of air means a fast moving fan meaning noise. So a good design would reject that. But if the heat sink is a good design, with adequate radiating area, that same CPU heat is transferred to a large area that even a small amount of moving air cools easily. And the heat sink does not get hot. If some of you would work with heat sinks this would be easy to understand.

No one's arguing those last 3 sentences.
You're making huge, absurd leaps of logic tying those last 3 sentences with the temp of the air coming out of a small exhaust vent. Your refusal to realize this is tiresome and trollsome.

It would be preferable if you would simply quote the sentences. "Small exhaust vent" - that indicates that the air is being funneled as it exits the case. This is bad design. If that's what you mean. If you have air going in at a certain rate and that airflow matches the air leaving (120mm in, 120mm out) there is no small exhaust vent. But if you had a 120mm in and a 60mm out there could be. But no intelligent design would do that. So I don't understand your wording.

I'm sure someone better at writing could have expressed this far more succinctly than I have. But anyone experimenting with heat sinks would understand. I keep on thinking of the human hand grasping a piece of metal for analogy. If I hold a cold, small piece of metal, that metal instantly warms up. Like a small heat sink. If I grasp a cold, large piece of metal, my hand may not change the temperature of the metal much at all. Like a large heat sink. In both cases warmth is leaving my hand and entering the metal and warming it. One gets much warmer than the other. That's why good heat sinks have large radiating areas. And that's undoubtedly why iMacs exhaust such hot air.

Message #15 - Posted 2007/06/07 - The New Guy

Previously, The New Guy wrote:

You hard-headedly refuse to admit that the temperature of the COMPONENTS is what really matters, not the temperature of exiting air!

One day it will dawn of some of you that if the temperature of a component rises and is not cooled with a heat sink, the temperature of the exiting air will be hot. The trick to this problem is to cool something with external air and by radiating the heat away from it as soon as possible. Or in the case of the hard drive, its the hard drive casing itself since we can't get air in there to cool the components that are heating the hard drive enclosure. So that's why a fan, using exterior ambient temperature air (not hot internal case air) works so admirably. A fan blowing on to it, not sucking hot air out of the case that just happens to wander past the hard drives.

There is another issue to consider. This has nothing to do with the original subject. Namely, the temperature of exiting air is directly related to the temperature of the room where the computer is located.

Yes - the 2 temperatures should be very close if the components are cooled well. As soon as something is getting hot, its a sign the heat sink is not up to the task. The same principle allows some heat sinks to cool CPU's with no fans at all. Just the slightest airflow is enough to whisk that heat away that is being radiated by all those cooling fins of the heat sink. And it follows that if one put a one of those large heat sinks, the air off that heat sink wouldn't be warm. And of course you'll notice every good heat sink these days uses heat pipes; almost all of them being expensive copper. So you can see why a manufacturer tries to shy away from expensive parts like that.

I have 2 macs in my rather large study that are on 24/7 and the room is always warmer than it should be. In the winter I close all the heating/AC floor registers and in the summer I open them + I have a ceiling fan. Of course, the hard disks contribute to the heat as well.

What model of Macs? It should be possible to replace the heat sinks in some models. Or by looking at the weakness of the design, one may rectify it.

For instance the G4 tower (specifically a dual 500 but others might be similar) can be cured so easily.
Disconnect the useless 4" fan.
Mount a 80 to 120 mm fan running at 5 volts (red and black wires off your Molex multicolored white plastic plugs that power things like DVD burners and IDE hard drives) so its silent, on the CPU heat sink. That will cool the CPU heat sink well. Now take apart the power supply and turn the fan so it blows in. Change the wires so its running at 5 volts (red and black). Now it will be almost inaudible. And it will be cooled much better now that you're not using hot interior case air to cool the power supply heat sink(s). The only downside to this scenario is you have to have the case open a crack to let the heat escape. But as long as you don't have a lot of dust, you're fine.

I forgot to mention that even though you're cooling the CPU heat sink with warmed case air (not warmed much thanks to the opening at the top for the heat to escape), this works better probably because the fan is closer to the heat sink. As in right on top.

I saw a G4 tower 1.25 ghz the other day and the weak part of this design was mounting the hard drives so the air passes by them before being conducted to the CPU heat sink. Great for the hard drives, bad for the CPU. It has provisions for 4 hard drives (maybe more) so just don't use the ones at the bottom.

Fans have gotten much quieter these days so sometimes just replacing an older fan with a newer one can reap huge benefits. Sleeve bearings are the worst, single ball bearings are better, dual ball bearings are best.

Message #16 - Posted 2007/06/07 - John McWilliams

The New Guy wrote:

Previously, The New Guy wrote:

Careless of me......didn't see that 7 drives were replaced (I'm guessing those were hard drives).

It's clear where the hot air is coming from.

lsmft

Message #17 - Posted 2007/06/08 - G.T.

The New Guy wrote:

You hard-headedly refuse to admit that the temperature of the COMPONENTS is what really matters, not the temperature of exiting air!

One day it will dawn of some of you that if the temperature of a component rises and is not cooled with a heat sink, the temperature of the exiting air will be hot.

For goodness sake. I haven't plonked anyone in 8 years but here you go, PLONK.

Greg

The ticketbastard Tax Tracker:
http://www.ticketmastersucks.org/tracker.html

Dethink to survive - Mclusky

Message #18 - Posted 2007/06/08 - Mike Rosenberg

The New Guy wrote:

One day it will dawn of some of you that if the temperature of a component rises and is not cooled with a heat sink, the temperature of the exiting air will be hot.

One day it will dawn on _YOU_, when you're no longer a teenager and no longer think you know everything, that just because you don't understand how something works doesn't mean it doesn't work. You're locked into this one concept of how it should work and fail to see there's a whole big world of things out there beyond your comprehension.

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://comedancing.biz> Ballroom dance shirts & gifts

Message #19 - Posted 2007/06/10 - The New Guy

And of course, if there is a small amount of air carrying off all of the heat, then the air gets hotter.

The air temperature will rise but if the heat sink has a large radiating area (fins) that rise will be minimal.

Wrong again. Those fins are not for radiation; they're for contact transfer of heat from the heat sink to the air.

Well I thought that was the term used. At least we're on the same page here. Radiating area is a term I've seen frequently used in the cooling industry, perhaps incorrectly. It usually means that a heat sink with a large radiating area will have an unusually high number of fins. And of course the heat sink itself will be large. The end result is that it cools better because its dissipating the heat better to the air. I think we agree on that point at least. Its pretty basic.

The surfaces of the fins face each other, for the most part; therefore, if they radiated heat, they would radiate to each other and keep the heat within the heat sink.

Well however you want to describe it, the fins transfer the heat to the air. Its doesn't really matter how it does it. But you will notice that all high performance heat sinks are large and have a high number of fins.

Since almost any fan gets noisy above 7 volts, that means you're not dealing with a lot of airflow to arrive at a cooling goal.

Voltage has little to do with the noise.

Voltage indicates the speed of the fan. Here your lack of hardware experience is showing. Almost any 80mm+ fan that works properly will be inaudible at 5 volts. Most will even in inaudible at 7 volts with the slightest amount of ambient noise to mask it. 12 volts usually results in noise. This may seem inconsequential, but if you're using multiple fans, its really important unless you don't mind working with a roaring computer. Which few of us wish to put up with.

So voltage is directly related to noise. I've never experienced an inaudible 120mm fan. To get them quiet, you need to reduce the rpm which means reducing the voltage.

Message #20 - Posted 2007/06/10 - The New Guy

the heat sink community

There's a "heat sink community"?
Man, get a life.

That wasn't the best choice of words. The cooling community would have been far better.

Message #21 - Posted 2007/06/10 - Mike Rosenberg

The New Guy wrote:

The cooling community would have been far better.

Either way, what the hell are you talking about? "The cooling community" would have to be defined by anyone but you as people who actually understand the concepts of thermodynamics and cooling.

YOUR community is rather small, consisting of you alone.

So, once again, how old are you and what are your educational credentials? (Do you actually believe that people are unable to see that you're ducking the question?)

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://designsbymike.info/shop/dance.cgi> Ballroom dance shirts & gifts

Message #22 - Posted 2007/06/10 - The New Guy

Here's the mindset of most people when it comes to cooling. They think mostly in terms of airflow.

That's right.

But noise standards precludes much airflow so its all on the shoulders of the heat sink.

All that a heat sink does is store heat. It's called a heat sink because it is to heat what a kitchen sink (or bathroom sink) is to water. It stores or holds it until it can be emptied. But just like a water sink with the water running, it keeps filling up unless a drain is opened. In the case of a heat sink, the air flow is analogous to the open drain.

Yes but remember that a better heat sink design transfers more heat to the air quicker. That's not a good choice of words but I'm tired and hopefully that will do. Better heat sinks move more heat away from the heat generator. There is a huge difference in the ability of heat sinks to move heat away from the heat generator. Some are far better than others and they are about the same size. The difference being in the heat pipe and fin design, as well as the materials used. Mac people seem to think only in adjusting the airflow (the speed in which the air moves or the fan rpm controlled by the changing the voltage), perhaps because Mac heat sinks are usually not user replaceable.

Its not so much about air movement - its about heat radiation area.

It has *nothing* to do with heat radiation. It has everything to do with heat conduction from the heat sink to the air, and then convection of the air out of the case. The purpose of the fins is to provide more contact area between the heat sink and the air.

By heat radiation I meant exactly that. If that's not the correct English word, I'm sorry.

The principle stands: as soon as you cool something with external air, exhaust that heated air immediately so it doesn't heat up anything else.

Exactly, and that is what the iMac does. It has nothing to do with radiation.

By radiation I meant how well that heat sink moves the heat away from the CPU. Should I have used the word convection? Either way, at least we're understanding each other better now. Unfortunately you don't seem to realize the huge difference in CONVECTION abilities of different heat sinks.

Message #23 - Posted 2007/06/10 - Michelle Steiner

Previously, The New Guy wrote:

There's a "heat sink community"? Man, get a life.

That wasn't the best choice of words. The cooling community would have been far better.

If you were a member of such a community (assuming that there actually is one), you would have been able to answer my question.

Support the troops: Bring them home ASAP.

Message #24 - Posted 2007/06/10 - Jolly Roger

On 2007-06-10 12:40:06 -0500, The New Guy said:

a heat
sink with a large radiating area will have an unusually high number of fins. And of course the heat sink itself will be large. The end result is that it cools better because its dissipating the heat better to the air. I think we agree on that point at least. Its pretty basic.

Larger heat sinks cool better only if air flow is constant, which it is NOT.

JR

Message #25 - Posted 2007/06/10 - Jolly Roger

On 2007-06-10 13:03:21 -0500, The New Guy said:

Mac
people seem to think only in adjusting the airflow (the speed in which the air moves or the fan rpm controlled by the changing the voltage)

No, it's just that *you* never think of air flow.

JR

Message #26 - Posted 2007/06/10 - Michelle Steiner

Previously, The New Guy wrote:

Wrong again. Those fins are not for radiation; they're for contact transfer of heat from the heat sink to the air.

Well I thought that was the term used.

You thought wrong--yet again.

At least we're on the same page here.

No we're not.

Radiating area is a term I've seen frequently used in the cooling industry, perhaps incorrectly.

Which goes to show that this mythical "cooling industry" doesn't know its own subject matter.

It usually means that a heat sink with a large radiating area will have an unusually high number of fins.

Heat sinks are not designed to radiate; the number of fins is related to contact area for conduction of heat; it has nothing to do with radiating heat.

The surfaces of the fins face each other, for the most part; therefore, if they radiated heat, they would radiate to each other and keep the heat within the heat sink.

Well however you want to describe it, the fins transfer the heat to the air.

A lot depends on how you describe it; you describe it incorrectly (as you did) and people know that you don't know what you're talking about.

But yes, the fins transfer the heat to the air. At least you finally understand that.

Its doesn't really matter how it does it.

Actually, it does matter how it does it.

But you will notice that all high performance heat sinks are large and have a high number of fins.

I noticed that long before you did, and I understand why they do. You have yet to show that you understand.

Since almost any fan gets noisy above 7 volts, that means you're not dealing with a lot of airflow to arrive at a cooling goal.

Voltage has little to do with the noise.

Voltage indicates the speed of the fan.

Not necessarily.

Here your lack of hardware experience is showing.

Not at all. You, on the other hand, have demonstrated your ignorance of hardware.

So voltage is directly related to noise.

Not at all.

The more you post, the more ignorance you display. There is nothing wrong with ignorance, but your refusal to learn makes you stupid.

Support the troops: Bring them home ASAP.

Message #27 - Posted 2007/06/10 - Michelle Steiner

Previously, The New Guy wrote:

It has *nothing* to do with heat radiation. It has everything to do with heat conduction from the heat sink to the air, and then convection of the air out of the case. The purpose of the fins is to provide more contact area between the heat sink and the air.

By heat radiation I meant exactly that.

Therefore you do not know what you're talking about.

If that's not the correct English word, I'm sorry.

You meant exactly what you said, but you don't know what you meant?

Exactly, and that is what the iMac does. It has nothing to do with radiation.

By radiation I meant how well that heat sink moves the heat away from the CPU.

Oh, I see; you think that you can redefine words to mean what you think they mean, regardless of their actual meanings.

That is not what radiation means, and no one who knows even an iota about the subject knows what radiation means.

You do not know anything about how computers and their components are cooled. If you happen to say anything that is correct on the subject, that would be strictly accidental.

Support the troops: Bring them home ASAP.

Message #28 - Posted 2007/06/10 - Mike Rosenberg

The New Guy wrote:

Mac people seem to think only in adjusting the airflow (the speed in which the air moves or the fan rpm controlled by the changing the voltage), perhaps because Mac heat sinks are usually not user replaceable.

We're a whole bunch of completely different people, of different ages and educational backgrounds who have one and one thing alone in common, that we use a Mac.

How old are you and what are your educational credentials?

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://designsbymike.info/shop/dance.cgi> Ballroom dance shirts & gifts

Message #29 - Posted 2007/06/10 - Mike Rosenberg

The New Guy wrote:

Well I thought that was the term used. At least we're on the same page here.

How old are you and what are your educational credentials?

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://designsbymike.info/shop/dance.cgi> Ballroom dance shirts & gifts

Message #30 - Posted 2007/06/10 - Mike Rosenberg

The New Guy wrote:

Here's the mindset of most people when it comes to cooling.

How old are you and what are your educational credentials?

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://designsbymike.info/shop/dance.cgi> Ballroom dance shirts & gifts

Message #31 - Posted 2007/06/10 - Jolly Roger

On 2007-06-10 13:35:59 -0500, Mike Rosenberg said:

The New Guy wrote:

Well I thought that was the term used. At least we're on the same page here.

How old are you and what are your educational credentials?

Mike, you know he's not going to answer. ; )

JR

Message #32 - Posted 2007/06/10 - Dave Balderstone

Previously, The New Guy wrote:

the heat sink community

There's a "heat sink community"?
Man, get a life.

That wasn't the best choice of words. The cooling community would have been far better.

How old are you and what are your educational credentials?

Message #33 - Posted 2007/06/10 - Dave Balderstone

Previously, The New Guy wrote:

By radiation I meant how well that heat sink moves the heat away from the CPU. Should I have used the word convection? Either way, at least we're understanding each other better now. Unfortunately you don't seem to realize the huge difference in CONVECTION abilities of different heat sinks.

How old are you and what are your educational credentials?

Message #34 - Posted 2007/06/10 - Mike Rosenberg

Jolly Roger wrote:

How old are you and what are your educational credentials?

Mike, you know he's not going to answer. ; )

Well, I've already said I'm assuming until he states otherwise that he's a 15-year-old smartass know-it-all like my nephew since he sounds just like my nephew. (And, for the record, Kenny is a great kid, I love him dearly, and he's extremely intelligent, but he also happens to be a 15-year-old smartass know-it-all.)

Besides, I have it on good authority that asking this question is catching on in the cooling thread community.

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://designsbymike.info/shop/dance.cgi> Ballroom dance shirts & gifts

Message #35 - Posted 2007/06/10 - Sue Rodgers

Jolly Roger wrote:

How old are you and what are your educational credentials?

Mike, you know he's not going to answer. ; )

Besides, I was not aware that talking out of one's ass actually requires credentials. ;-)

Living is easy with eyes closed, misunderstanding all you see...

Message #36 - Posted 2007/06/10 - Tom Stiller

Previously, The New Guy wrote:

By radiation I meant how well that heat sink moves the heat away from the CPU. Should I have used the word convection? Either way, at least we're understanding each other better now. Unfortunately you don't seem to realize the huge difference in CONVECTION abilities of different heat sinks.

There are three modes of heat transfer:
1. Conduction, in which the bodies must be in physical contact (e.g. CPU to heatsink),
2. Convection, in which a heat conducted from the hot body to a fluid (e.g. air) is subsequently carried by natural means (e.g. motion due to density changes in the fluid) or forced (e.g. a fan or other pump),
3. Radiation, which requires no transport medium (i.e. can take place in a vacuum) and depends on the nature of the surface and its temperature.

So you should have used the term "conduction". By themselves, heatsinks have no "convection ability". The effectiveness of a particular scheme depends on the chosen fluid and its flow over the surfaces of the heatsink.

Tom Stiller

PGP fingerprint = 5108 DDB2 9761 EDE5 E7E3
7BDA 71ED 6496 99C0 C7CF

Message #37 - Posted 2007/06/10 - Adrian

Sue Rodgers wrote:

Jolly Roger wrote:

How old are you and what are your educational credentials?

Mike, you know he's not going to answer. ; )

Besides, I was not aware that talking out of one's ass actually requires credentials. ;-)

Yes, it does. You have to be a member of the heat sink community.

Adrian

Message #38 - Posted 2007/06/10 - Dave Balderstone

Previously, Adrian wrote:

Sue Rodgers wrote:

Jolly Roger wrote:

How old are you and what are your educational credentials?

Mike, you know he's not going to answer. ; )

Besides, I was not aware that talking out of one's ass actually requires credentials. ;-)

Yes, it does. You have to be a member of the heat sink community.

Oh, there's SO much fun to be had with that exchange...

But I shall refrain, as dinner needs to be prepared.

Message #39 - Posted 2007/06/10 - Mike Rosenberg

Dave Balderstone wrote:

Oh, there's SO much fun to be had with that exchange...

But I shall refrain, as dinner needs to be prepared.

Come on, Dave, you _could_ order a pizza...

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://designsbymike.info/shop/dance.cgi> Ballroom dance shirts & gifts

Message #40 - Posted 2007/06/10 - Dave Balderstone

Previously, Mike Rosenberg wrote:

Dave Balderstone wrote:

Oh, there's SO much fun to be had with that exchange...

But I shall refrain, as dinner needs to be prepared.

Come on, Dave, you _could_ order a pizza...

Rather than grilled chicken, steamed asparagus, and homemade potato salad? Sirrah, I think NOT.

But you have caught me out. Dinner was only an excuse. I refrained from following the logical direction solely out of respect for the members of this group who have somewhat delicate sensibilities.

I just didn't want to brag about it...

Message #41 - Posted 2007/06/10 - Mike Rosenberg

Dave Balderstone wrote:

Rather than grilled chicken, steamed asparagus, and homemade potato salad? Sirrah, I think NOT.

I don't care much for aspargus but you had me with the grilled chicken.

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://designsbymike.info/shop/dance.cgi> Ballroom dance shirts & gifts

Message #42 - Posted 2007/06/10 - Michelle Steiner

Previously, Dave Balderstone wrote:

Yes, it does. You have to be a member of the heat sink community.

Oh, there's SO much fun to be had with that exchange...

Is there an overlap between the heat sink community and time traveling industrial rocker terrorists? We need a Venn diagram.

Enjoy your dinner.

Support the troops: Bring them home ASAP.

Message #43 - Posted 2007/06/10 - The New Guy

Radiation area is a term used in the heat sink community to denote amount of space used to radiate accumulated heat. Just because a heat sink is large, doesn't necessarily mean it has a large radiation area. Its all to do with the fins used. Newer, good designs have many finely made fins. The latest generation Scythe Ninja is pretty much regarded as king in this department. Its a good example of a large radiation area.

A large surface area, yes. Not a large *radiation* area.

I used "radiation" because the surface area transfers the heat to the air. Heaters radiate heat - it seems similar.

"heat sink community" ... good grief!

That wasn't the best choice of words. I need to proof read more. Though here its used:
http://www.hexus.net/content/news/news_archive_month.php?month=200202 "Lately, there has been a new trend evolving in the heatsink community: all the big mofo's are not only growing in size, they all come with 80mm fans too. A few months ago, all the powerful heatsinks were using high RPM fans, resulting in whining rigs that are really disturbing to the user and his environment (how many folks can leave their computer up and running while they sleep with a Delta fan whining ... I know I can't!). Since installing even faster fans was not an option, there had to come a new way to improve cooling capabilities, and that is by using larger heatsinks with larger fans."

That was written in 2002. Very true today.

Message #44 - Posted 2007/06/10 - The New Guy

My whole point was that the temperature of the exiting air is extremely indicative of the cooling design. That by just holding your hand in front of the exhaust you can instantly tell things - as long as the computer is not doing some resource demanding tasks

But you agreed with my scenario of the 2 computers with different sized heatsinks inside expelling the same amount of heat ... so holding your hand in front of the exhaust doesn't help you.

So to maintain the same temperature the airflow has to be regulated in one of them to match the temperature in the other one.
The problem is that is not the way you regulate temperature in a good design. You run the fan at the highest inaudible speed. Then you use a heat sink that dissipates enough heat to keep things cool. You're depending on airflow for cooling. Airflow (due to noise restraints) is very limited. You need to depend on heat sink design instead.

All I know is that if you used a massive heat sink it wouldn't even get warm. The small amount of air movement would pull any rise in temperature away with the result that exiting air would be near the ambient temperature. Maybe some of you are making this a lot more complicated than it need be.
1 - Cool with external air.
2 - Exhaust heated air immediately.
The only way to do this that I can tell is for motherboards to be designed for each heat generator (CPU, Video card, chipset, ram) to be mounted on the corner. Otherwise its hard to exhaust the air without that heated air affecting something else. As components get hotter and hotter, this will become more and more critical. Designers have been able to get away with cooling with warm air for a long time. But won't work for much longer. The Mac Pro is a step in that direction. A very compromised design, mind you, but its a step. Cooling the RAM heat sinks with hot CPU heated air doesn't seem very logical.

Here's a product designed to lower the temps of the Mac Pro's ram: http://www.maxupgrades.com/istore/index.cfm?fuseaction=product.display& Product_ID=160
But the silly thing is they are not addressing the source of the air. Its coming right from the CPU heat sinks, the largest heat generator in the computer. Too bad they couldn't duct the CPU heat out the side of the case, then pull in the RAM cooling air from the bottom ducting it out the back like it presently is. If you're going to go to all the trouble to put huge heat sinks on the ECC FD-Dimms, doesn't it make sense not to blow hot air on them?

Maybe in the next version of the Mac Pro they will have positive air cooling on the hard drives (all 4 instead of sort of cooling the middle 2 only) instead of cramming them upside down so the heat is trapped (hard drives radiate far more heat from the bottom than the top). And have some of them hooked up with eSata so they are hot swappable and accessible from the front. That would be great for people working with large HD video files. To have to open the case to change a hard drive (not to mention having to shut the system down, is just wrong. I'm sure eSata is on its way. 3 gb/sec plus hot swappability is great for everybody. Sure beats firewire. Too bad Apple has such a close relationship with firewire. It might take longer than it should.

Message #45 - Posted 2007/06/10 - Michelle Steiner

Previously, The New Guy wrote:

I used "radiation" because the surface area transfers the heat to the air.

You used "radiation" because you don't know anything about the topic.

Support the troops: Bring them home ASAP.

Message #46 - Posted 2007/06/10 - The New Guy

I noticed that nobody has commented on the 25C temperature of the positively cooled example mentioned on the SPCR website. And no, there wasn't a discrepancy between the C or F or whatever. They were both in C. 25C vs 49C. If 49C is typical, be aware that your hard drive is going to die way before it needs to.

No one commented because there was nothing to comment about. Obviouly you can alter the cooling scenario to give greater cooling to a given component.

The point was that Mac cultists think that 49C is good cooling for hard drives. When an example is sited showing how to halve that using something no more complicated than a simple fan, they ignore it. Go figure. The real point is when figures show that Apple has done a lousy job of something so very simple, they can't admit it. Its all quiet on the home front in Appleland.

Your contention is that that a drive running at 25C will last significantly longer than one routinely running at higher temps. People have certainly made comments on that.

Most any electronics tech will admit that heat is not your friend. So if getting it cool is not so difficult, how stupid or stubborn (is there a difference?) do you have to be to not take advantage of such an easy solution?

Maybe it's a small factor, I can't say, but it's clearly not a big one as we would all have found out by now from our own experiences .... oh and how about:

I doubt that because so few systems use active hard drive cooling.

"Google doubts hard drives fail because of excessive temperature, usage" "In the end, Google's research does not solve the problem of predicting when hard drives are likely to fail. However, it shows that temperature and high usage alone are not responsible for failures by default." http://www.tgdaily.com/content/view/30990/113/

Its saying there are other factors. Duh.

I'm saying if heat is a factor and it can be solved, do it!

100,000 drives is quite a study environment! Here's an excerpt: "Failures do not increase when the average temperature increases. In fact, there is a clear trend showing that lower temperatures are associated with higher failure rates. Only at very high temperatures is there a slight reversal of this trend,"

But they don't mention any temperatures? Maybe they are all run way higher than a well cooled system. So they are just comparing warm to hot drives instead of cool to everything else.

An interesting point is the relevancy of scan errors. I wonder if heat affects scan errors. Because apparently, according to Google, scan errors just get worse. Sort of like a hard drive cancer.

It also says that temps and high usage are not the sole reason for drive errors. Well I guess we knew that already. I hope Google does better work on their software than their hard drive data!

Message #47 - Posted 2007/06/10 - The New Guy

"Google doubts hard drives fail because of excessive temperature, usage"

"In the end, Google's research does not solve the problem of predicting when hard drives are likely to fail. However, it shows that temperature and high usage alone are not responsible for failures by default."

http://www.tgdaily.com/content/view/30990/113/

Dude, don't bother. The New Guy doesn't care about real-world data or facts. He lives in a delusion where he, and only he, knows all.

Real world data should have mentioned temperatures as the article discusses heat. This is strangely absent. Maybe the study is available in greater detail elsewhere with the temps mentioned.

Message #48 - Posted 2007/06/10 - G.T.

Dave Balderstone wrote:

Previously, Mike Rosenberg wrote:

Dave Balderstone wrote:

Oh, there's SO much fun to be had with that exchange...

But I shall refrain, as dinner needs to be prepared.

Come on, Dave, you _could_ order a pizza...

Rather than grilled chicken, steamed asparagus, and homemade potato salad? Sirrah, I think NOT.

My Dad's homegrown asparagus cooked on the grill. Damn nice. Since I don't own a grill currently I steam at home but I look forward to my parents' or my girlfriend's especially when we have some fresh asparagus.

And there's nothing like ice cold homemade potato salad on a hot summer day whichever way it's been prepared.

Greg

The ticketbastard Tax Tracker:
http://www.ticketmastersucks.org/tracker.html

Dethink to survive - Mclusky

Message #49 - Posted 2007/06/11 - Adrian

The New Guy wrote:

So to maintain the same temperature the airflow has to be regulated in one of them to match the temperature in the other one.
The problem is that is not the way you regulate temperature in a good design. You run the fan at the highest inaudible speed. Then you use a heat sink that dissipates enough heat to keep things cool. You're depending on airflow for cooling. Airflow (due to noise restraints) is very limited. You need to depend on heat sink design instead.

Yes, the cooling in my example depends on the smaller heat sink having higher airflow. But ALL heat sinks need airflow. That airfflow is usually provided by use of a fan. A well designed case makes best use of the airflow provided by the fan by channelling the air, and of course uses quiet fans. Of course you can have a large box with everything using large surface area heat sinks and rely purely on natural convection currents to take away the excess heat (no fans) ... but there's a choice being made by the manufacturer, and each method can be equally effective or ineffective. If you want to design your case to be space efficient you can't always rely on natural convection (though apple did specifically use this in some of its earlier designs).

Adrian

Message #50 - Posted 2007/06/11 - Adrian

The New Guy wrote:

All I know is that if you used a massive heat sink it wouldn't even get warm. The small amount of air movement would pull any rise in temperature away with the result that exiting air would be near the ambient temperature. Maybe some of you are making this a lot more complicated than it need be.

That's where you are getting confused. If you had a aeht sink the size of the Earth it would still be warmed by the hot component. A heat sink with a large surface area can pass its heat to the air more effectively than one with a small surface area, *all other things being equal*. However, it's just the same amount of heat spread over a bigger area .... so the amount of heat which has to be removed, and which emerges from the case is the same as for a small heat sink with higher airflow. Look up the First Law of Thermodynamics on Wikipedia. This is really basic stuff. You MUST have learned this at school if you did any physics at all.

Adrian

Message #51 - Posted 2007/06/11 - The New Guy

The cooling community would have been far better.

Either way, what the hell are you talking about? "The cooling community" would have to be defined by anyone but you as people who actually understand the concepts of thermodynamics and cooling.

All the people that make heat sinks, fans, liquid cooling, etc. That should have been obvious.

So, once again, how old are you and what are your educational credentials? (Do you actually believe that people are unable to see that you're ducking the question?)

No matter what I answer you will use it against me. What's the point?

Message #52 - Posted 2007/06/11 - The New Guy

There's a "heat sink community"?

That wasn't the best choice of words. The cooling community would have been far better.

If you were a member of such a community (assuming that there actually is one), you would have been able to answer my question.

I never implied I am a member. Just someone that has worked with fans and heat sinks and have noticed what works and what doesn't. And how much noise it took to accomplish whatever you're aiming for.

Message #53 - Posted 2007/06/11 - The New Guy

So to maintain the same temperature the airflow has to be regulated in one of them to match the temperature in the other one.
The problem is that is not the way you regulate temperature in a good design. You run the fan at the highest inaudible speed. Then you use a heat sink that dissipates enough heat to keep things cool. You're depending on airflow for cooling. Airflow (due to noise restraints) is very limited. You need to depend on heat sink design instead.

Yes, the cooling in my example depends on the smaller heat sink having higher airflow. But ALL heat sinks need airflow. That airfflow is usually provided by use of a fan.

99% of all computers use fans and heat sinks. That we all agree on.

A well designed case makes best use of
the airflow provided by the fan by channelling the air, and of course uses quiet fans. Of course you can have a large box with everything using large surface area heat sinks and rely purely on natural convection currents to take away the excess heat (no fans) ...

I don't think that any of us are discussing computers with no fans. Not really applicable.

Message #54 - Posted 2007/06/11 - The New Guy

All I know is that if you used a massive heat sink it wouldn't even get warm. The small amount of air movement would pull any rise in temperature away with the result that exiting air would be near the ambient temperature. Maybe some of you are making this a lot more complicated than it need be.

That's where you are getting confused. If you had a heat sink the size of the Earth it would still be warmed by the hot component.

A well designed heat sink (usually large) will get less hot. A large heat sink might get just slightly above ambient room temperature. Therefore the exiting air will be slightly above ambient room temperature. And that's what I've been saying from the start.

A heat sink
with a large surface area can pass its heat to the air more effectively than one with a small surface area, *all other things being equal*. However, it's just the same amount of heat spread over a bigger area .... so the amount of heat which has to be removed, and which emerges from the case is the same as for a small heat sink with higher airflow. Look up the First Law of Thermodynamics on Wikipedia. This is really basic stuff. You MUST have learned this at school if you did any physics at all.

Well I was very poor in many aspects of physics and excelled in other aspects of physics so maybe this is a weak area of mine. I'm having a hard time relating what you just said to computer case cooling. Maybe its because the larger heat sink doesn't need nearly as much air flow to stay cool? In the real world, you slap a better heat sink on there and it runs cooler. So you can turn your fan down and have the same temperature. Or keep the fan the same and have a lower temperature. Its pretty simple. If you've worked with fans and heat sinks this would be plain and simple. My original point which most everyone has missed is that Apple's problem is its choice of heat sinks and the use of warmed interior air to cool them. The Mac Pro seems to be the only design that addresses (and only with the CPU heat sinks I think). In a typical PC motherboard layout, if one cooled the CPU heat sink from the side panel (exterior air), then vented it straight up and out (after removing the power supply), then did the same for the video card and vented that straight up and slightly towards the middle, then mounted the optical and hard drives vertically and put a wall between the front area and the motherboard with a fan on the bottom blowing air up (with no horizontal drives to block vertical air movement) you just might have a great design. Hideous perhaps, but functional.

Message #55 - Posted 2007/06/11 - The New Guy

By radiation I meant how well that heat sink moves the heat away from the CPU. Should I have used the word convection? Either way, at least we're understanding each other better now. Unfortunately you don't seem to realize the huge difference in CONVECTION abilities of different heat sinks.

There are three modes of heat transfer:
1. Conduction, in which the bodies must be in physical contact (e.g. CPU to heatsink),
2. Convection, in which a heat conducted from the hot body to a fluid (e.g. air) is subsequently carried by natural means (e.g. motion due to density changes in the fluid) or forced (e.g. a fan or other pump),
3. Radiation, which requires no transport medium (i.e. can take place in a vacuum) and depends on the nature of the surface and its temperature.

So you should have used the term "conduction". By themselves, heatsinks have no "convection ability". The effectiveness of a particular scheme depends on the chosen fluid and its flow over the surfaces of the heatsink.

Good explanation.

Now check this out.
http://ieeexplore.ieee.org/Xplore/login.jsp?url=/iel6/8217/25229/011351 14.pdf
"An accurate straightforward technique is presented to compute thermal radiation from finned heat sinks. A set of graphs is included to aid in the computational procedure. The analytical approach presented compares favorably with experimental data."
Should they have used "thermal conduction"? I've always understood that heat radiation is used to describe a heated surface heating the air. Like a heat sink does.

Another one here:
http://www.engineersedge.com/heat_transfer/extruded-heat-sink-radiation .htm
"Use the following equation to calculate the radiation from an extruded (finned) heat sink."

Googling "heatsink radiation" brings tons of hits. They seem to be talking about the ability of the heat sink to transfer heat from the heat generator to the air so the fan can take it away.

Message #56 - Posted 2007/06/11 - The New Guy

Wrong again. Those fins are not for radiation; they're for contact transfer of heat from the heat sink to the air.

Well I thought that was the term used.

In another post I show a couple of websites that use the term radiation to describe the ability of the heat sink to move heat from the heat generator to the air. Its quite widely used. Google "heatsink radiation" and you'll see.

Radiating area is a term I've seen frequently used in the cooling industry, perhaps incorrectly.

Which goes to show that this mythical "cooling industry" doesn't know its own subject matter.

Well they're in the business - you'd think they would use accurate words to describe their product's function.

It usually means that a heat sink with a large radiating area will have an unusually high number of fins.

Heat sinks are not designed to radiate; the number of fins is related to contact area for conduction of heat; it has nothing to do with radiating heat.

You must have been a top notch debater in school, Miehelle. But content here is more important.

Aren't we getting bogged down with semantics? Would anyone disagree that the heat sink's function is to move heat (conduct, whatever) from the CPU to the air? The fins get warm, that warmth warms the air, and the fans move that air away so it doesn't build up. Pretty simple, no?

The surfaces of the fins face each other, for the most part; therefore, if they radiated heat, they would radiate to each other and keep the heat within the heat sink.

Well however you want to describe it, the fins transfer the heat to the air.

A lot depends on how you describe it; you describe it incorrectly (as you did) and people know that you don't know what you're talking about.

But yes, the fins transfer the heat to the air. At least you finally understand that.

I think a retarded chimpanzee with Alzheimer's would understand that. Semantics, semantics.
The point is Apple's heat sinks are far from their potential. And their hard drives run way hotter than they need to.

Since almost any fan gets noisy above 7 volts, that means you're not dealing with a lot of airflow to arrive at a cooling goal.

Voltage has little to do with the noise.

Voltage indicates the speed of the fan.

Not necessarily.

Show me one fan controller that doesn't use voltage to change the speed of the fan.

So voltage is directly related to noise.

Not at all.

Oh really? I don't know about your part of the world, but here in mine, when voltage is changed the fan speed changes. That translates to a difference of noise. Is it really different in Planet Michelle?

The more you post, the more ignorance you display. There is nothing wrong with ignorance, but your refusal to learn makes you stupid.

Have you ever even touched a fan? Ever hooked up a fan to 5 volts, 7 volts and 12 volts? If you had, you'd see the speed changes with the noise of course changing. Why are we even talking about such mundane issues?

Message #57 - Posted 2007/06/11 - The New Guy

Mac people seem to think only in adjusting the airflow (the speed in which the air moves or the fan rpm controlled by the changing the voltage)

No, it's just that *you* never think of air flow.

Thats because in a well designed system there is not a lot of variability in airflow. In a lousy design, if it gets hotter the fan starts howling, like in my Mini sometimes. But it should be engineered for inaudibility under normal operating conditions at the lowest possible temperatures. Some of you (that occasionally peer out of your closed Mac box) may notice that sub 120 mm fans are not used much at all these days, because of their noise and lack of airflow. You'll notice that in the Mac Pro compared to the G5 tower. Big fans are going to really change the computer case world. Already 200 and 250 mm fans are becoming more and more common. Now if you run a large fan at 5 volts it will be inaudible. And you'll have more airflow. But an 80 mm type fan moves so little air most designs insist on 12 volt operating voltages and that makes them very noisy.

So I do think of airflow. A lot.

Message #58 - Posted 2007/06/11 - The New Guy

a heat
sink with a large radiating area will have an unusually high number of fins. And of course the heat sink itself will be large. The end result is that it cools better because its dissipating the heat better to the air. I think we agree on that point at least. Its pretty basic.

Larger heat sinks cool better only if air flow is constant, which it is NOT.

Larger heat sinks cool better because they transfer (conduct?) heat from the heat generator to the air better.
"Only if air flow is constant"?????
I don't follow that at all. When is airflow constant in any computer? Maybe 5 or more years ago.
In my experience a better designed heat sink requires less airflow to stay at the same temperature than a lousy designed heat sink. Run a few 120+mm fans at 5 volts and you have inaudible airflow. Then engineer the heat sink to cool whatever you want with that available airflow. The result is a silent computer. Under very occasional high workloads the fans should go faster but that is very infrequent. Otherwise its just another noisy computer.

Message #59 - Posted 2007/06/11 - Andy

Mike Rosenberg wrote:

The New Guy wrote:

Here's the mindset of most people when it comes to cooling.

How old are you and what are your educational credentials?

I'm 112 and I've been heavily involved with several Heat Sink Communities (we just like to call 'em HSCs).

Message #60 - Posted 2007/06/11 - Tom Stiller

Previously, The New Guy wrote:

I've always understood
that heat radiation is used to describe a heated surface heating the air. Like a heat sink does.

Well, you're mistaken. Thermal radiation is the mechanism by which the sun warms the earth, no air and no direct contact.

A heat sink heats the air (directly) by conduction and the warmed air is moved away from the source (to be replaced by cooler air) by convection, either natural or forced.

Tom Stiller

PGP fingerprint = 5108 DDB2 9761 EDE5 E7E3
7BDA 71ED 6496 99C0 C7CF

Message #61 - Posted 2007/06/11 - Mike Rosenberg

The New Guy wrote:

So, once again, how old are you and what are your educational credentials? (Do you actually believe that people are unable to see that you're ducking the question?)

No matter what I answer you will use it against me. What's the point?

We wouldn't use it against you if you actually had any relevant credentials and weren't just a smartass know-it-all 15-year-old. When you're a few years older, you'll get to the age where you actually understand how obnoxious people your age can be in their smug fantasy world in which they think they see all, know all, smell all.

I'm not saying you're not intelligent, just that you're typical of kids your age in talking out your ass and thinking you're infallible.

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://designsbymike.info/shop/dance.cgi> Ballroom dance shirts & gifts

Message #62 - Posted 2007/06/11 - Michelle Steiner

Previously, The New Guy wrote:

Aren't we getting bogged down with semantics?

No, we're getting bogged down with your ignorance and arrogance.

Support the troops: Bring them home ASAP.

Message #63 - Posted 2007/06/11 - Tim Streater

Previously, The New Guy wrote:

Wrong again. Those fins are not for radiation; they're for contact transfer of heat from the heat sink to the air.

Well I thought that was the term used.

The fins are for convective cooling of the heat sink. Just like on your radiator in your car, and your house.

In another post I show a couple of websites that use the term radiation to describe the ability of the heat sink to move heat from the heat generator to the air. Its quite widely used. Google "heatsink radiation" and you'll see.

Radiating area is a term I've seen frequently used in the cooling industry, perhaps incorrectly.

Since the heatsink is painted black, it will be a reasonable radiator if its hot - that is, it will radiate heat, like any hot body (in fact, any body at all radiates heat, even if its a block of ice). But all that does is get absorbed by components in the vicinity. What you want is to heat the air (and so cool down the heatsink) by *convective* heat transfer.

Perhaps we better explain what convection is. The air in direct contact with the hot object gets heated by conduction, just as your finger is when it touches something hot. This hotter air expands and moves away, allowing more cooler air to come in contact with the hot object. The moving away is the convective part, and is assisted by a fan.

Note that a fan has no effect *at* *all* on the radiative heat loss from the heatsink.

Radiative area is a reasonable term if that's how you're trying to achieve radiative cooling. The fins on a heatsink are helpful for convective cooling as they increase the total surface area, but are less useful for radiative cooling since you've got bits of heatsink radiating at other bits of heatsink.

Notice that a hot yellow gas flame (say, methane) *radiates* a lot of heat, whereas a burning hydrogen flame does not. So you can stand next to a large hydrogen flame but not a large methane one. Don't stand *over* either, though :-)

Which goes to show that this mythical "cooling industry" doesn't know its own subject matter.

Well they're in the business - you'd think they would use accurate words to describe their product's function.

It usually means that a heat sink with a large radiating area will have an unusually high number of fins.

Heat sinks are not designed to radiate; the number of fins is related to contact area for conduction of heat; it has nothing to do with radiating heat.

You must have been a top notch debater in school, Miehelle. But content here is more important.

Aren't we getting bogged down with semantics? Would anyone disagree that the heat sink's function is to move heat (conduct, whatever) from the CPU to the air? The fins get warm, that warmth warms the air, and the fans move that air away so it doesn't build up. Pretty simple, no?

The surfaces of the fins face each other, for the most part; therefore, if they radiated heat, they would radiate to each other and keep the heat within the heat sink.

Well however you want to describe it, the fins transfer the heat to the air.

A lot depends on how you describe it; you describe it incorrectly (as you did) and people know that you don't know what you're talking about.

But yes, the fins transfer the heat to the air. At least you finally understand that.

Michelle's right about this - in fact I didn't read this bit until I finished writing mine, above, but notice what I wrote matches what she wrote.

Sure, the heatsink gets rid of heat. But if you don't understand the three methods of heat transfer (conduction, radiation, convection), and what is happening in each case, you won't really understand how a heatsink works.

Message #64 - Posted 2007/06/11 - Jolly Roger

On 2007-06-11 01:48:17 -0500, The New Guy said:

There's a "heat sink community"?

That wasn't the best choice of words. The cooling community would have been far better.

If you were a member of such a community (assuming that there actually is one), you would have been able to answer my question.

I never implied I am a member. Just someone that has worked with fans and heat sinks and have noticed what works and what doesn't. And how much noise it took to accomplish whatever you're aiming for.

Just curious: How much experience do you have with systems that have all of these?:

* completely (or nearly-completely) separate component air chambers

* built-in component and ambient temperature sensors within the case

* firmware and software that automatically adjust air flow rates based on sensed temperatures

JR

Message #65 - Posted 2007/06/11 - G.T.

Jolly Roger wrote:

On 2007-06-11 01:48:17 -0500, The New Guy said:

There's a "heat sink community"?

That wasn't the best choice of words. The cooling community would have been far better.

If you were a member of such a community (assuming that there actually is one), you would have been able to answer my question.

I never implied I am a member. Just someone that has worked with fans and heat sinks and have noticed what works and what doesn't. And how much noise it took to accomplish whatever you're aiming for.

Just curious: How much experience do you have with systems that have all of these?:

* completely (or nearly-completely) separate component air chambers

* built-in component and ambient temperature sensors within the case

* firmware and software that automatically adjust air flow rates based on sensed temperatures

Cool. My filter is working. Now I'll just learn a little more about thermodynamics from the bright people in c.s.m.s. like Stiller, Streater, and Steiner (not that those are the brightest but I like the sound of the three St names) and then I'll kill the whole thread. Funny that the New Guy doesn't want to learn from these smart folks and Adrian, Balderstone, BreadWithSpam, Mitchell, and Rosenburg, and funnier that THERE IS NOT A SINGLE PERSON HERE supporting the New Guy's arguments.

Greg

The ticketbastard Tax Tracker:
http://www.ticketmastersucks.org/tracker.html

Dethink to survive - Mclusky

Message #66 - Posted 2007/06/11 - Michelle Steiner

Previously, G.T. wrote:

funnier that THERE IS NOT A SINGLE PERSON HERE supporting the New Guy's arguments.

That is because New Guy is an expert and knows everything there is to know about the subject, and the rest of us don't know anything.

Support the troops: Bring them home ASAP.

Message #67 - Posted 2007/06/11 - Adrian

The New Guy wrote:

A well designed heat sink (usually large) will get less hot. A large heat sink might get just slightly above ambient room temperature. Therefore the exiting air will be slightly above ambient room temperature. And that's what I've been saying from the start.

Yes, of course the heat will be spread over a larger area, so the temperature on any one part of that large heat sink will be lower than any one part of a smaller heat sink ... all other things being equal. Believe it or not nobody has been arguing with that!

The point is that the heat must be transferred to the air and exit the case. The part you are missing is that the total amount of heat which emerges from the case when a component is successfully cooled and maintained at a given temperature will be the same, no matter what size the heat sink.

Let me turn it around. You can heat your room with a 1 bar electric fire running at a given temperature or you can use a 2 bar fire with each of the elements at half the temperature. The room will b heated to the same temperature. (You may notice tha a large surface area central heating radiator will heat your room just as much as a small intense ehat source like an electric fire. This is analagous to your heat sinks. A small heat sink at double the temperature will heat the case the same amount as a double sized one at half the temperature. So the air coming out of the case has to carry the same heat. Think of your heat sinks as minature electric fires ... do you get it?

Adrian

Message #68 - Posted 2007/06/11 - Adrian

The New Guy wrote:

I don't think that any of us are discussing computers with no fans. Not really applicable.

Actually, there was a version of the see-through iMac DV which which had no fans! It even had a heat-producing cathode ray tube in there! Supposedly they were designed to remove heat entirely by natural convection. It was a brave move ... whether they all died an early heat death I couldn't tell you, but I never heard that they did 1

Adrian

Message #69 - Posted 2007/06/11 - Adrian

The New Guy wrote:

I used "radiation" because the surface area transfers the heat to the air. Heaters radiate heat - it seems similar.

Yes, it's a commonly accepted use of language which is actually technically wrong (I remember my old physics teacher speaking about this). We call central heating panels "radiators" but they pass the great majority of the heat by convection (that's why they have a large surface area, a little like a heat sink). They would be more accurately called convectors ... though we all still call them radiators. True radiators are heat sources that you can feel halfway across your room ... like a bright the bright glowing elements on old style electric fires.

Adrian

Message #70 - Posted 2007/06/11 - Adrian

The New Guy wrote:

The point was that Mac cultists think that 49C is good cooling for hard drives. When an example is sited showing how to halve that using something no more complicated than a simple fan, they ignore it. Go figure. The real point is when figures show that Apple has done a lousy job of something so very simple, they can't admit it. Its all quiet on the home front in Appleland.

What has "Mac cultists" got to do with anything? If the temperature is well within the spec for running a hard drive then what's the problem? Halving the temperature doesn't double the lifespan! In fact one of the articles I read (nothing to do with Apple) related to the Google experience of running thousands of drives pointed out that though there was no evidence of a relationship between high temperatures and drive failure there *was* a relationship between low temperatures and drive failure (they quoted less than 68F as low).

Would you try to run your car engine at 25C? Believe me, you had better not!

Adrian

Message #71 - Posted 2007/06/11 - Adrian

The New Guy wrote:

"Google doubts hard drives fail because of excessive temperature, usage"

"In the end, Google's research does not solve the problem of predicting when hard drives are likely to fail. However, it shows that temperature and high usage alone are not responsible for failures by default."

http://www.tgdaily.com/content/view/30990/113/

Dude, don't bother. The New Guy doesn't care about real-world data or facts. He lives in a delusion where he, and only he, knows all.

Real world data should have mentioned temperatures as the article discusses heat. This is strangely absent. Maybe the study is available in greater detail elsewhere with the temps mentioned.

"Strangely absent" you cheeky bugger! Just because you're too lazy to find the article yourself. The whole article is available with all the data present. I would have thought that with your special interest in the topic you would easily have tracked it down and absorbed its contents!

Its title is Failure Trends in a Large Disk Drive Population by Eduardo Pinheiro, Wolf-Dietrich Weber and Luiz Andre Barroso of Google Inc.

If you really can't find it I'll post even post a URL if you ask nicely!

Adrian

Message #72 - Posted 2007/06/11 - Michelle Steiner

Previously, Adrian wrote:

I don't think that any of us are discussing computers with no fans. Not really applicable.

Actually, there was a version of the see-through iMac DV which which had no fans! It even had a heat-producing cathode ray tube in there!

All the G3 CRT iMacs were fanless.

Supposedly they were designed to remove heat entirely by natural convection. It was a brave move ...

Not all that brave, the original Macintosh, through at least the Mac Plus (and maybe the SE as well) were also fanless.

Support the troops: Bring them home ASAP.

Message #73 - Posted 2007/06/11 - Michelle Steiner

Previously, Adrian wrote:

The part you are missing is that the total amount of heat which emerges from the case when a component is successfully cooled and maintained at a given temperature will be the same, no matter what size the heat sink.

The problem is that New Guy doesn't understand, and apparently refuses to understand, the difference between heat and temperature.

Support the troops: Bring them home ASAP.

Message #74 - Posted 2007/06/11 - Tom Stiller

Previously, Adrian wrote:

The New Guy wrote:

I don't think that any of us are discussing computers with no fans. Not really applicable.

Actually, there was a version of the see-through iMac DV which which had no fans! It even had a heat-producing cathode ray tube in there! Supposedly they were designed to remove heat entirely by natural convection. It was a brave move ... whether they all died an early heat death I couldn't tell you, but I never heard that they did 1

As I sit here looking at my (nearly) six-year old, convection-cooled iMac DV, which is on 24/7, I can assure that at least *one* of them is still alive.

Tom Stiller

PGP fingerprint = 5108 DDB2 9761 EDE5 E7E3
7BDA 71ED 6496 99C0 C7CF

Message #75 - Posted 2007/06/11 - Adrian

The New Guy wrote:

Larger heat sinks cool better because they transfer (conduct?) heat from the heat generator to the air better.
"Only if air flow is constant"?????
I don't follow that at all. When is airflow constant in any computer? Maybe 5 or more years ago.
In my experience a better designed heat sink requires less airflow to stay at the same temperature than a lousy designed heat sink. Run a few 120+mm fans at 5 volts and you have inaudible airflow. Then engineer the heat sink to cool whatever you want with that available airflow. The result is a silent computer. Under very occasional high workloads the fans should go faster but that is very infrequent. Otherwise its just another noisy computer.

Let me agree with you: Larger, properly designed heat sinks do indeed conduct heat from the heat source and then pass it on to the air more effectively.

Your latter description sounds something like my G5: "Under very occasional high workloads the fans should go faster but that is very infrequent."

I do have a cooling story for you. My Powermac was supplied with a video card (Radeon 9600XT) which was effectively cooled without an onboard fan ... just a heat sink taking advantage of the airflow through the case itself. I subsequently upgraded to a Radeon 9800 Pro 256MB. (I wanted best frame rates for X-Plane.) After I installed it I could have wept. My beautifully quiet computer had a constant whine from the fan on the new card ... I ascertained that everything was working as it should but there was always a noticeable whine in the background, rising in voume when under demand. A search on the web showed that others had the same experience. I could not put up with it and bought an aftermarket Verax heat sink and fan unit from Germany ... ah, blessed relief (and a thinner wallet!)

I give you this story simply to let you know that I do not put up with poor performance, noisy computers or inadequate cooling! If something is not right I'll say it, and do something about it!

Adrian

Message #76 - Posted 2007/06/11 - The New Guy

So, once again, how old are you and what are your educational credentials? (Do you actually believe that people are unable to see that you're ducking the question?)

No matter what I answer you will use it against me. What's the point?

We wouldn't use it against you if you actually had any relevant credentials and weren't just a smartass know-it-all 15-year-old.

That brings up an interesting point. Credentials don't cool a computer. Good design does. I've seen my fill of pompous, educated people, filled with credentials that have very poor results. But my, do they ever have a pile of credentials. I look at results - that's what matters in the long run. And often, new ideas are squashed by people filled to the brim with credentials. Look only at the message, never the messenger.

When you're a few years older, you'll get to the age where you actually understand how obnoxious people your age can be in their smug fantasy world in which they think they see all, know all, smell all. I'm not saying you're not intelligent,

LOL.....lots of others have. You don't have to. Doesn't matter. I've read about many brilliant people whose innovative ideas were scoffed and derided. In fact, most of the major inventions and new ideas were not accepted initially. They often took decades to be allowed into the mainstream. There's a saying I love: "Don't worry about people stealing your idea. If its any good, you'll have to ram it down people's throats." Its a testament to our innate stubbornness as a society. Of all the people that have responded, very, very few have anything consequential to say. Surprising as most of you are really intelligent and eloquent. But I bring up and dash some sacred cows and that's gotta hurt a bit. Also the lack of hardware experience the vast majority of Mac users have is undoubtedly playing a part here. I deal with lots of different users and talking to the typical Mac person about hardware is like talking to a child. They have all sorts of irrational fears and insecurities. My point is that this doesn't have to be so. But Apple loves this as it sells more computers. The more naive the customer the higher the profit. Time to buy more Apple stock.

just that you're typical of kids
your age in talking out your ass and thinking you're infallible.

Well then you've not been reading my posts. I frequently admit errors and shortcomings. Especially in the software department, where Michelle presently claims dominion over us lowly newbies. And rightly so. They have helped me so often. That's why I can't get angry when they are rude. I'm indebted.

Message #77 - Posted 2007/06/11 - The New Guy

The part you are missing is that the total amount of heat which emerges from the case when a component is successfully cooled and maintained at a given temperature will be the same, no matter what size the heat sink.

The problem is that New Guy doesn't understand, and apparently refuses to understand, the difference between heat and temperature.

Please explain as you see it.

Some of you people are analyzing this to death. Its so simple. You're getting caught up with semantics and definitions. The point is to cool your computer in a silent fashion. You're getting way off topic. I have very cool running computers that are dead quiet 99% of the time. I've never seen a Mac do that. Even the hallowed Mac Pro. So I guess I prove my point. What I say works because I've seen it. My components exhaust hear ambient temperature 99% of the time. That's the goal and I achieved it. It started with the 500mhz G4 tower and then the Mini as well as various PC's over the years. Cool with exterior air. Exhaust immediately. Everything else is a compromise.

Message #78 - Posted 2007/06/11 - Adrian

Tom Stiller wrote:

Previously, Adrian wrote:

The New Guy wrote:

I don't think that any of us are discussing computers with no fans. Not really applicable.

Actually, there was a version of the see-through iMac DV which which had no fans! It even had a heat-producing cathode ray tube in there! Supposedly they were designed to remove heat entirely by natural convection. It was a brave move ... whether they all died an early heat death I couldn't tell you, but I never heard that they did 1

As I sit here looking at my (nearly) six-year old, convection-cooled iMac DV, which is on 24/7, I can assure that at least *one* of them is still alive.

Point proven! Apple must know something about designing computers which are properly cooled after all ... funny that!

Adrian

Message #79 - Posted 2007/06/11 - Adrian

Michelle Steiner wrote:

Previously, Adrian wrote:

I don't think that any of us are discussing computers with no fans. Not really applicable.

Actually, there was a version of the see-through iMac DV which which had no fans! It even had a heat-producing cathode ray tube in there!

All the G3 CRT iMacs were fanless.

Supposedly they were designed to remove heat entirely by natural convection. It was a brave move ...

Not all that brave, the original Macintosh, through at least the Mac Plus (and maybe the SE as well) were also fanless.

True. (Mind you, the amount of heat being produced is quite a bit different ... but with a CRT in the case they had to know what they were doing!)

Adrian

Message #80 - Posted 2007/06/11 - Adrian

The New Guy wrote:

I've always understood
that heat radiation is used to describe a heated surface heating the air. Like a heat sink does.

... and you would be wrong. That's convection:

http://en.wikipedia.org/wiki/Convection

Adrian

Message #81 - Posted 2007/06/11 - Adrian

Andy wrote:

Mike Rosenberg wrote:

The New Guy wrote:

Here's the mindset of most people when it comes to cooling.

How old are you and what are your educational credentials?

I'm 112 and I've been heavily involved with several Heat Sink Communities (we just like to call 'em HSCs).

I like that! Being a member of an HSC ... kind of gives you a warm cosy feeling of belonging.

Adrian

Message #82 - Posted 2007/06/11 - Michelle Steiner

Previously, The New Guy wrote:

Credentials don't cool a computer. Good design does.

Not that you know anything about good design.

Well then you've not been reading my posts. I frequently admit errors and shortcomings. Especially in the software department, where Michelle presently claims dominion over us lowly newbies.

Liar! I never claimed any such thing.

And the only times you have ever admitted any errors is when you haven't been able to bluff your way out of it.

Support the troops: Bring them home ASAP.

Message #83 - Posted 2007/06/11 - Michelle Steiner

Previously, Adrian wrote:

I like that! Being a member of an HSC ... kind of gives you a warm cosy feeling of belonging.

If the heat sink community were properly designed, it would give you a cool feeling. That warm feeling means it's a poor design; find another HSC.

Support the troops: Bring them home ASAP.

Message #84 - Posted 2007/06/11 - Michelle Steiner

Previously, The New Guy wrote:

The problem is that New Guy doesn't understand, and apparently refuses to understand, the difference between heat and temperature.

Please explain as you see it.

It has been explained to you many times. And if you understood it, you wouldn't have had any problems with the question I had posed to you.

Some of you people are analyzing this to death. Its so simple.

It is simple, but you don't understand it.

You're getting caught up with semantics and definitions.

We're not getting caught up with semantics and definitions; we're expounding basic principles.

Support the troops: Bring them home ASAP.

Message #85 - Posted 2007/06/11 - Adrian

The New Guy wrote:

Some of you people are analyzing this to death. Its so simple. You're getting caught up with semantics and definitions. The point is to cool your computer in a silent fashion. You're getting way off topic.

Back to the topic then. We are all arguing that you putting your hand where warm air exits a computer and then spouting off about poor design is evident nonsense.

Adrian

Message #86 - Posted 2007/06/11 - Adrian

Michelle Steiner wrote:

Previously, Adrian wrote:

I like that! Being a member of an HSC ... kind of gives you a warm cosy feeling of belonging.

If the heat sink community were properly designed, it would give you a cool feeling. That warm feeling means it's a poor design; find another HSC.

You are just a Mac cultist anyway. Going on about your cool HSC. My HSC has a big radiator so it's bound to be cooler than yours.

Adrian

Message #87 - Posted 2007/06/11 - Mike Rosenberg

Adrian wrote:

You are just a Mac cultist anyway. Going on about your cool HSC. My HSC has a big radiator so it's bound to be cooler than yours.

I keep all my equipment immersed in liquid nitrogen, but I'm envious of the liquid helium community.

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://designsbymike.info/shop/dance.cgi> Ballroom dance shirts & gifts

Message #88 - Posted 2007/06/11 - Michelle Steiner

Previously, Adrian wrote:

I like that! Being a member of an HSC ... kind of gives you a warm cosy feeling of belonging.

If the heat sink community were properly designed, it would give you a cool feeling. That warm feeling means it's a poor design; find another HSC.

You are just a Mac cultist anyway. Going on about your cool HSC. My HSC has a big radiator so it's bound to be cooler than yours.

But if your HSC is cool, how can it give you a warm feeling?

Support the troops: Bring them home ASAP.

Message #89 - Posted 2007/06/11 - Michelle Steiner

Previously, Mike Rosenberg wrote:

You are just a Mac cultist anyway. Going on about your cool HSC. My HSC has a big radiator so it's bound to be cooler than yours.

I keep all my equipment immersed in liquid nitrogen, but I'm envious of the liquid helium community.

Go for liquid hydrogen.

Support the troops: Bring them home ASAP.

Message #90 - Posted 2007/06/11 - Mike Rosenberg

Michelle Steiner wrote:

I keep all my equipment immersed in liquid nitrogen, but I'm envious of the liquid helium community.

Go for liquid hydrogen.

Well, I'd like to be able to use the gaseous helium to fill balloons and to talk funny.

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://designsbymike.info/shop/dance.cgi> Ballroom dance shirts & gifts

Message #91 - Posted 2007/06/11 - Michelle Steiner

Previously, Mike Rosenberg wrote:

I keep all my equipment immersed in liquid nitrogen, but I'm envious of the liquid helium community.

Go for liquid hydrogen.

Well, I'd like to be able to use the gaseous helium to fill balloons and to talk funny.

You talk funny anyway. Besides, with the gaseous hydrogen, you can build a working model of the Hindenberg in your back yard.

Support the troops: Bring them home ASAP.

Message #92 - Posted 2007/06/11 - Mike Rosenberg

Michelle Steiner wrote:

Well, I'd like to be able to use the gaseous helium to fill balloons and to talk funny.

You talk funny anyway.

I walked right into that one, didn't I?

Besides, with the gaseous hydrogen, you can build a working model of the Hindenberg in your back yard.

That's what I'm afraid of. Oh, the humanity!

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://designsbymike.info/shop/dance.cgi> Ballroom dance shirts & gifts

Message #93 - Posted 2007/06/11 - Joey DoWop Dee

On Jun 11, 2007, Michelle Steiner wrote:

Previously, Adrian wrote:

I like that! Being a member of an HSC ... kind of gives you a warm cosy feeling of belonging.

If the heat sink community were properly designed, it would give you a cool feeling. That warm feeling means it's a poor design; find another HSC.

You are just a Mac cultist anyway. Going on about your cool HSC. My HSC has a big radiator so it's bound to be cooler than yours.

But if your HSC is cool, how can it give you a warm feeling?

It's by Microsoft.

Joey DoWop Dee
Remember: It is To Laugh

Message #94 - Posted 2007/06/11 - Tom Stiller

Previously, Michelle Steiner wrote:

Previously, Mike Rosenberg wrote:

You are just a Mac cultist anyway. Going on about your cool HSC. My HSC has a big radiator so it's bound to be cooler than yours.

I keep all my equipment immersed in liquid nitrogen, but I'm envious of the liquid helium community.

Go for liquid hydrogen.

Has anyone studies the behavior of semiconductors at temperatures below 20 K?

Tom Stiller

PGP fingerprint = 5108 DDB2 9761 EDE5 E7E3
7BDA 71ED 6496 99C0 C7CF

Message #95 - Posted 2007/06/11 - Mike Rosenberg

Tom Stiller wrote:

Has anyone studies the behavior of semiconductors at temperatures below 20 K?

I haven't. I've been too busy practicing a ballroom dance routine lately.

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://designsbymike.info/shop/dance.cgi> Ballroom dance shirts & gifts

Message #96 - Posted 2007/06/11 - The New Guy

Previously, Adrian wrote:

The New Guy wrote:

"Google doubts hard drives fail because of excessive temperature, usage"

"In the end, Google's research does not solve the problem of predicting when hard drives are likely to fail. However, it shows that temperature and high usage alone are not responsible for failures by default."

http://www.tgdaily.com/content/view/30990/113/

Dude, don't bother. The New Guy doesn't care about real-world data or facts. He lives in a delusion where he, and only he, knows all.

Real world data should have mentioned temperatures as the article discusses heat. This is strangely absent. Maybe the study is available in greater detail elsewhere with the temps mentioned.

"Strangely absent" you cheeky bugger! Just because you're too lazy to find the article yourself. The whole article is available with all the data present. I would have thought that with your special interest in the topic you would easily have tracked it down and absorbed its contents!
Its title is Failure Trends in a Large Disk Drive Population by Eduardo Pinheiro, Wolf-Dietrich Weber and Luiz Andre Barroso of Google Inc. If you really can't find it I'll post even post a URL if you ask nicely!

If I was writing an article about fast cars the primary ingredients of that article would be the speed those cars attained. Is that logical? If you're writing about temperatures and computers it seems odd that one wouldn't mention a temperature anywhere in the article. I wasn't trying to cheeky or a bugger. Its simply relevant and logical. Especially as it wouldn't take up a lot of space in the article. Maybe one sentence.

Found the article right away. Thanks.

"Ä Contrary to previously reported results, we found very little correlation between failure rates and either elevated temperature or activity levels."

Since over 99% of computers in general use are poorly cooled, I wonder what percentage of Google's hard drives were in the 15 to 30C range. And how do you even get a hard drive down to 15C? That's way below any room temperature. They don't talk about that.

Interesting results nevertheless.

Message #97 - Posted 2007/06/11 - The New Guy

The point was that Mac cultists think that 49C is good cooling for hard drives. When an example is sited showing how to halve that using something no more complicated than a simple fan, they ignore it. Go figure. The real point is when figures show that Apple has done a lousy job of something so very simple, they can't admit it. Its all quiet on the home front in Appleland.

What has "Mac cultists" got to do with anything? If the temperature is well within the spec for running a hard drive then what's the problem? Halving the temperature doesn't double the lifespan! In fact one of the articles I read (nothing to do with Apple) related to the Google experience of running thousands of drives pointed out that though there was no evidence of a relationship between high temperatures and drive failure there *was* a relationship between low temperatures and drive failure (they quoted less than 68F as low).

Actually they said 15C - to 30C. That's quite a spread. In fact its way too much of a spread for much of a conclusion. Still they are convinced and you have to respect that since they perhaps have the best test bench in the world thanks to the number of drives involved. Strange why they didn't put more emphasis on temperature since that is one environmental aspect largely in control of the user whereas some of the others simply are not.

Perhaps running moving parts at less than typical room temperature is not so great - but not one in a thousand people is going to do that. Virtually all hard drives are above typical room temperature.

I sure wish they had been more specific on those temperatures. Its an interesting read, that's for sure.

Message #98 - Posted 2007/06/11 - Adrian

The New Guy wrote:

Since over 99% of computers in general use are poorly cooled,

Are they? OK stick to your principles ... give us the evidence. --
Adrian

Message #99 - Posted 2007/06/12 - Adrian

Michelle Steiner wrote:

Previously, Adrian wrote:

I like that! Being a member of an HSC ... kind of gives you a warm cosy feeling of belonging.

If the heat sink community were properly designed, it would give you a cool feeling. That warm feeling means it's a poor design; find another HSC.

You are just a Mac cultist anyway. Going on about your cool HSC. My HSC has a big radiator so it's bound to be cooler than yours.

But if your HSC is cool, how can it give you a warm feeling?

It's that smug sense of satisfaction I feel, rather than a temperature. I've got lovely fins.

Adrian

Message #100 - Posted 2007/06/11 - Jim Gibson

Previously, Michelle Steiner wrote:

Previously, Adrian wrote:

I don't think that any of us are discussing computers with no fans. Not really applicable.

Actually, there was a version of the see-through iMac DV which which had no fans! It even had a heat-producing cathode ray tube in there!

All the G3 CRT iMacs were fanless.

Supposedly they were designed to remove heat entirely by natural convection. It was a brave move ...

Not all that brave, the original Macintosh, through at least the Mac Plus (and maybe the SE as well) were also fanless.

The Mac Cube is also fanless.

Posted Via Usenet.com Premium Usenet Newsgroup Services ---------------------------------------------------------- ** SPEED ** RETENTION ** COMPLETION ** ANONYMITY ** ---------------------------------------------------------- http://www.usenet.com

Message #101 - Posted 2007/06/11 - Michelle Steiner

Previously, Jim Gibson wrote:

The Mac Cube is also fanless.

Oh, I'm sure that it has quite a number of fans.

Support the troops: Bring them home ASAP.

Message #102 - Posted 2007/06/11 - Michelle Steiner

Previously, Adrian wrote:

I like that! Being a member of an HSC ... kind of gives you a warm cosy feeling of belonging.

If the heat sink community were properly designed, it would give you a cool feeling. That warm feeling means it's a poor design; find another HSC.

You are just a Mac cultist anyway. Going on about your cool HSC. My HSC has a big radiator so it's bound to be cooler than yours.

But if your HSC is cool, how can it give you a warm feeling?

It's that smug sense of satisfaction I feel, rather than a temperature. I've got lovely fins.

You're a porpoise?

Support the troops: Bring them home ASAP.

Message #103 - Posted 2007/06/11 - Michelle Steiner

Previously, Tom Stiller wrote:

Has anyone studies the behavior of semiconductors at temperatures below 20 K?

Do you expect us to do your research for you? Google it yourself, mister! muhaaahaahh.

Support the troops: Bring them home ASAP.

Message #104 - Posted 2007/06/11 - Michelle Steiner

Previously, Mike Rosenberg wrote:

Has anyone studies the behavior of semiconductors at temperatures below 20 K?

I haven't. I've been too busy practicing a ballroom dance routine lately.

Ballroom dance routine? Did you have adequate cooling, or did you sweat?

Support the troops: Bring them home ASAP.

Message #105 - Posted 2007/06/11 - Michelle Steiner

Previously, Mike Rosenberg wrote:

Well, I'd like to be able to use the gaseous helium to fill balloons and to talk funny.

You talk funny anyway.

I walked right into that one, didn't I?

You surely did.

Besides, with the gaseous hydrogen, you can build a working model of the Hindenberg in your back yard.

That's what I'm afraid of. Oh, the humanity!

<http://tinyurl.com/2nhfap>

Support the troops: Bring them home ASAP.

Message #106 - Posted 2007/06/11 - Mike Rosenberg

Michelle Steiner wrote:

The Mac Cube is also fanless.

Oh, I'm sure that it has quite a number of fans.

Yeah, but not Giant fans.

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://designsbymike.info/shop/dance.cgi> Ballroom dance shirts & gifts

Message #107 - Posted 2007/06/11 - Mike Rosenberg

Michelle Steiner wrote:

I haven't. I've been too busy practicing a ballroom dance routine lately.

Ballroom dance routine? Did you have adequate cooling, or did you sweat?

It's a foxtrot, so it's more important to LOOK cool than to BE cool. And for smooth dances, we _perspire_; in Latin dances we sweat.

<http://designsbymike.info/shop/mac.cgi> Mac and geek T-shirts & gifts <http://designsbymike.info/shop/musings.cgi> Muckraking T-shirts <http://designsbymike.info/shop/prius.cgi> Prius shirts/bumper stickers <http://designsbymike.info/shop/dance.cgi> Ballroom dance shirts & gifts

Message #108 - Posted 2007/06/11 - Matthew T. Russotto

Previously, The New Guy wrote:

lowest possible temperatures. Some of you (that occasionally peer out of your closed Mac box) may notice that sub 120 mm fans are not used much at all these days, because of their noise and lack of airflow.

Horsepucky. There are plenty of 80mm case fans, most CPU fans are 80mm, power supply fans are largely 80mm, and 92mm case fans are still far more common than 120mm.

You'll notice that in the Mac Pro compared to the G5 tower. Big fans are going to really change the computer case world. Already 200 and 250 mm fans are becoming more and more common.

A fan almost a foot in diameter is common? No, I don't think so. --
There's no such thing as a free lunch, but certain accounting practices can result in a fully-depreciated one.

Message #109 - Posted 2007/06/11 - Matthew T. Russotto

Previously, Adrian wrote:

Michelle Steiner wrote:

Previously, Adrian wrote:

I like that! Being a member of an HSC ... kind of gives you a warm cosy feeling of belonging.

If the heat sink community were properly designed, it would give you a cool feeling. That warm feeling means it's a poor design; find another HSC.

You are just a Mac cultist anyway. Going on about your cool HSC. My HSC has a big radiator so it's bound to be cooler than yours.

Oh yeah? I'm a member of the heat SOURCE community. Darn ambiguous acronyms.

There's no such thing as a free lunch, but certain accounting practices can result in a fully-depreciated one.

Message #110 - Posted 2007/06/11 - Matthew T. Russotto

Previously, The New Guy wrote:

Since over 99% of computers in general use are poorly cooled, I wonder what percentage of Google's hard drives were in the 15 to 30C range. And how do you even get a hard drive down to 15C? That's way below any room temperature. They don't talk about that.

Most likely they cool the ambient air with a split system using phase change refrigeration technology. That would be pretty exotic stuff... if this were the first half of the 20th century instead of the 21st. As it is, most Americans probably have them nowadays.

There's no such thing as a free lunch, but certain accounting practices can result in a fully-depreciated one.

Message #111 - Posted 2007/06/11 - Michelle Steiner

Previously, Mike Rosenberg wrote:

The Mac Cube is also fanless.

Oh, I'm sure that it has quite a number of fans.

Yeah, but not Giant fans.

If you're in San Francisco or in New York, there's a good chance that you're a Giants fan.

Support the troops: Bring them home ASAP.

Message #112 - Posted 2007/06/11 - Michelle Steiner

Previously, Mike Rosenberg wrote:

Ballroom dance routine? Did you have adequate cooling, or did you sweat?

It's a foxtrot, so it's more important to LOOK cool than to BE cool. And for smooth dances, we _perspire_; in Latin dances we sweat.

So you think you can dance? I didn't see you on TV.

Support the troops: Bring them home ASAP.

Message #113 - Posted 2007/06/12 - Adrian

Michelle Steiner wrote:

Previously, Adrian wrote:

I like that! Being a member of an HSC ... kind of gives you a warm cosy feeling of belonging.

If the heat sink community were properly designed, it would give you a cool feeling. That warm feeling means it's a poor design; find another HSC.

You are just a Mac cultist anyway. Going on about your cool HSC. My HSC has a big radiator so it's bound to be cooler than yours.

But if your HSC is cool, how can it give you a warm feeling?

It's that smug sense of satisfaction I feel, rather than a temperature. I've got lovely fins.

You're a porpoise?

That's the whole porpoise, cooling.

Adrian

Message #114 - Posted 2007/06/12 - Adrian

Matthew T. Russotto wrote:

Previously, Adrian wrote:

Michelle Steiner wrote:

Previously, Adrian wrote:

I like that! Being a member of an HSC ... kind of gives you a warm cosy feeling of belonging.

If the heat sink community were properly designed, it would give you a cool feeling. That warm feeling means it's a poor design; find another HSC.

You are just a Mac cultist anyway. Going on about your cool HSC. My HSC has a big radiator so it's bound to be cooler than yours.

Oh yeah? I'm a member of the heat SOURCE community. Darn ambiguous acronyms.

Damn you, Thermo Man, I'll drain every last kelvin from you ... --
Adrian

Message #115 - Posted 2007/06/12 - The New Guy

Previously, Adrian wrote:

The New Guy wrote:

Larger heat sinks cool better because they transfer (conduct?) heat from the heat generator to the air better.
"Only if air flow is constant"?????
I don't follow that at all. When is airflow constant in any computer? Maybe 5 or more years ago.
In my experience a better designed heat sink requires less airflow to stay at the same temperature than a lousy designed heat sink. Run a few 120+mm fans at 5 volts and you have inaudible airflow. Then engineer the heat sink to cool whatever you want with that available airflow. The result is a silent computer. Under very occasional high workloads the fans should go faster but that is very infrequent. Otherwise its just another noisy computer.

Let me agree with you: Larger, properly designed heat sinks do indeed conduct heat from the heat source and then pass it on to the air more effectively.

Your latter description sounds something like my G5: "Under very occasional high workloads the fans should go faster but that is very infrequent."

I do have a cooling story for you. My Powermac was supplied with a video card (Radeon 9600XT) which was effectively cooled without an onboard fan ... just a heat sink taking advantage of the airflow through the case itself. I subsequently upgraded to a Radeon 9800 Pro 256MB. (I wanted best frame rates for X-Plane.) After I installed it I could have wept. My beautifully quiet computer had a constant whine from the fan on the new card ... I ascertained that everything was working as it should but there was always a noticeable whine in the background, rising in voume when under demand. A search on the web showed that others had the same experience. I could not put up with it and bought an aftermarket Verax heat sink and fan unit from Germany ... ah, blessed relief (and a thinner wallet!)

I give you this story simply to let you know that I do not put up with poor performance, noisy computers or inadequate cooling! If something is not right I'll say it, and do something about it!

Interesting story. Video cards, because of their smaller size of heat sinks generally are much more challenging to cool simply because its harder to fit a larger fan on there. And small fans are usually noisy. I haven't worked with any high end cards yet so I have no input there.

Message #116 - Posted 2007/06/12 - The New Guy

Previously, Michelle Steiner wrote:

Previously, The New Guy wrote:

Well then you've not been reading my posts. I frequently admit errors and shortcomings. Especially in the software department, where Michelle presently claims dominion over us lowly newbies.

Liar! I never claimed any such thing.

Wow - we're getting our panties in a little bit of knot, aren't we.... It was a light hearted expression. Sorry the tone was lost on your interpretation of it. And it was also meant as a compliment to your knowledge in that area.

Message #117 - Posted 2007/06/12 - The New Guy

Previously, Matthew T. Russotto wrote:

Previously, The New Guy wrote:

lowest possible temperatures. Some of you (that occasionally peer out of your closed Mac box) may notice that sub 120 mm fans are not used much at all these days, because of their noise and lack of airflow.

Horsepucky. There are plenty of 80mm case fans, most CPU fans are 80mm, power supply fans are largely 80mm, and 92mm case fans are still far more common than 120mm.

You're looking at old technology which is being quickly supplanted. Look at any overclocking rag and you'll see very few, if any sub 120mm fans. Especially case fans. Tigerdirect sells 5 packs for about $25. That's pretty cheap!

You'll notice that in the Mac Pro compared to the G5 tower. Big fans are going to really change the computer case world. Already 200 and 250 mm fans are becoming more and more common.

A fan almost a foot in diameter is common? No, I don't think so.

They are getting more and more common. The Antec 900 case uses a 200 mm fan and Thermaltake uses a 250 mm fan. There are others but these are large case manufacturers which use them on their higher end models. So its clear where the industry is heading. When you think about it, it makes sense. Why use something that moves x amount of air at 40 db when you can use something else that moves the same amount of air at 25 db? Finally most computer users are realizing that they don't have to put up with it anymore (the noise.) Coupled with careful mounting systems and a lack of air restriction from gratings that are too close to the blades (causes a sort of whooshing sound) and you really get progress in the quiet department.

Message #118 - Posted 2007/06/12 - The New Guy

Previously, Matthew T. Russotto wrote:

Previously, The New Guy wrote:

Since over 99% of computers in general use are poorly cooled, I wonder what percentage of Google's hard drives were in the 15 to 30C range. And how do you even get a hard drive down to 15C? That's way below any room temperature. They don't talk about that.

Most likely they cool the ambient air with a split system using phase change refrigeration technology. That would be pretty exotic stuff... if this were the first half of the 20th century instead of the 21st. As it is, most Americans probably have them nowadays.

So a server room with very aggressive air conditioning? I sure would have loved to have known what percentage of the units were kept that cool. Must be expensive. I bet it was a sliver of the total.

Need Help? Have a Question?

Looking for more help, comments, and answers?

Ask your questions on Ask Different. Ask Different is a community of Apple users ready to help.