Heat dumping?

[j_eckel]

In this thread Edward Ng said something that jumped out at me:

...with or without a fan on its cooler, that graphics card will dump the same amount of heat into your case. The only situation where an active cooler would reduce ramping of other fans in the system is an active cooler that actually exhausts heat out of the system, like an Arctic Cooling Silencer.

Maybe this is basic stuff to everyone but me, but I've always thought that passive cooling dumped more heat into the case than a cooler with a fan. For example, a Zalman VF700 would keep the VGA cooler than a Zalman ZM80D-HP could, and therefore dump less heat into the case.

Am I misunderstanding something here?

[wooglin]

I think what they mean is that unless the video cooler is actually exhausting heat to the outside of the case, whether its cooling is fan or passive it will dump the same amount of heat inside the case.

A cooler with a fan (non exhausting) is the same as a cooler without a fan.

[j_eckel]

I think what they mean is that unless the video cooler is actually exhausting heat to the outside of the case, whether its cooling is fan or passive it will dump the same amount of heat inside the case.

A cooler with a fan (non exhausting) is the same as a cooler without a fan.

This assumes that passive and active keep the VGA at the same temp, right? I can't understand how a passive cooler that keeps the VGA at 55c idle would dump the same amount of heat inside the case as a active cooler that keeps the VGA at 37c. *confused*

[swivelguy2]

I can't understand how a passive cooler that keeps the VGA at 55c idle would dump the same amount of heat inside the case as a active cooler that keeps the VGA at 37c. *confused*

The amount of heat dumped into the case is equal the power consumed by the device, which depends only on how much it is being used.

A passive VGA cooler will be hotter, and so the air it dumps into the case will be very hot, but without a fan, this will be a small volume of air. An active VGA cooler will keep the core cooler by circulating more air past the fins, so that air will not be as hot, but there will be LOTS of warm air.

So the difference between passive and active is the difference between producing a little very hot air, or much warm air, and it turns out to the same heat production for the two coolers.

The advantage of an active cooler is that the core is indeed at a low temp, so there's less of a chance of crashes and better overclocking potential. The advantage of a passive cooler is that there's no noise or vibration.

An active cooler that exhausts the air outside the case is often the best solution, because it adds to total case airflow while not dumping extra heat inside the case, so everything stays cooler.

Hope that helps make things clear.

edit: lol, the word "solution" turns into an ad link for some obscure MS office thing

edit 2: haha, wumpus, I win.

[wumpus]

The same amount of heat is generated no matter WHAT you use to cool the card.. active, passive, or even a banana.* Physics at work! First law of thermodynamics: energy is never created or destroyed, it can only be converted to different forms.

So, the heat is the same, it's just a question of where the heat goes, and how fast.

With a passive solution, there's not any ADDITIONAL heat, it's just more localized-- which is why you see it as "hotter". The heat doesn't move as far as it would if you had a fan actively blowing air across it.

With an active solution, the heat is moved away from the card faster, so the case (and the room.. and the world..) heat up but the card appears to be "cooler".

* not recommended

[Zyzzyx]

I can't understand how a passive cooler that keeps the VGA at 55c idle would dump the same amount of heat inside the case as a active cooler that keeps the VGA at 37c. *confused*

What you're confusing is heat and temperature.

The video card will maintain a constant (relatively) amount of heat output, rated in watts. Say something around 50watts of heat. This is heat produced by the video card and will always be transferred to the inside of a case (unless exhausted as with an AC cooler).

That one cooler will have the card at 55c and another at 37c is just a measure of how well those coolers remove that 50watts of heat from the video card. The cooler at 37c is more effective than the cooler at 55c. Regardless, they are both working with 50watts of heat.

[j_eckel]

swivelguy2, wumpus and Zyzzyx,

big thanks for taking the time to explain this.

As you've probably already have figured out, I'm considering passive cooling of my videocard, but I'm afraid that my PSU-fan will speed up and that I'd have to turn up the fanspeed of my CPU a notch or two due to a (possibly) warmer case.

Would passive cooling of the VGA lead to higher speeds of the other fans?

[Tobias]

Depends. It is a question of which active cooling. If we are talking about stock cooling, then no. A passive heatsink will not cause the other fans to spin up more than the stock cooling. (actually, since the fan will produce some tenth of a watt or something, the stock cooling will cause more heat than passive cooling)

The key is to look at how much heat the other fans have to coop with. The GPU will produce say 50 watts no matter what cooling and with either of the cooling solutions I've talked about use the air inside the case to cool down. The sum of it is that the amount of energy in the case is the same with eirther cooling solution. The difference is that is that the active cooling is more efficient due to the amount of air passing over the heatsink (compare with the windchill factor).

As you have understood there is a third option. Arctic cooling nv-silencer. This cooler blows all the heated air outside the case and by doing this reduce the amount of energy in the case, which lets the other fans (if temp-controlled) to be more effective (cooling a hotter surface and/or cooling with cooler air is more efficient) and may work less.

Of course, this has its price. Another fan with a direct soundpath out of the case, at that. But if that fan runs at noicelevels below ambient, or if you can not heat it for other reasons, that doesn't matter.

On the other hand, if you are concidering passive cooling you probably are not concidering any highpowered GPU (that is, say ATI9600XXX or lower) and those GPU:s doesn't produce much heat anyway and with a goop CPU-cooler and a 120mm PSU, that extra heat would not cause any signifikant extra trouble for cooling. Even very little airflow makes wonders if you have a good heatsink.

[EricTerminator]

Hi !

I'd like to give you an example of heat dissipation ; I have an idle human being and a Pentium 4 CPU, with Prescott core.

Both dissipates the same ammount of heat, rated in Watts.

But, because of physics, the first one can be passively cooled, while the other would melt. The CPU would reach an awfully hot temperature. It won't melt if you actively cool it, e.g. if you use a fan.

But if we were in a tight room, if I put the human being inside it, or the CPU inside it, the room would be heated to the same temperature.

Imagine that the room is the PC case, that the CPU is an actively cooled graphic card, the human being an passively cooled one. Same result ; the temperature in the case will be the same.

In fact, you thought that a passively cooled card dumps more heat in the case than an actively one because you hit the radiators, which are always hot.

But it's again the same case than the one of the CPU and the human ; it's just a question of surface. If the radiators were as big as the surface of the Earth, would you suspect that the CPU dumps as much heat as you ?

As another guy explained to you, the heat is just differently distributed to the air, but the amount of heat remains the same.

See you