Does water cooling contribute less to a room's ambient temperature?

[amsga]

Hi all,

This might sound like a stupid question but I'm interested to know if liquid cooling your PC contributes less heat compared to air cooling to the room.

Thanks!

 

[Supermuncher85]

No. It would violate the first law of thermodynamics. Since I can't be asked to type it all out myself

The first law of thermodynamics states that energy cannot be created or destroyed; rather, the amount of energy lost in a steady state process cannot be greater than the amount of energy gained. This is the statement of conservation of energy for a thermodynamic system. It refers to the two ways that a closed system transfers energy to and from its surroundings – by the processes of heat and mechanical work. The rate of gain or loss in the stored energy of a system is determined by the rates of these two processes. In open systems, the flow of matter is another energy transfer mechanism, and extra terms must be included in the expression of the first law.

http://en.wikipedia.org/wiki/Laws_of_thermodynamics#First_law

Energy is given off to the radiator, which in turn has to be cooled by fans. Therefore the total amount of heat given off will be proportional to the power dissipation off the CPU/GPU respectively. This will not change, regardless of what type of cooling you use.
The advantages however are:
1.) Can use larger quiet fans to cool radiator compared to what CPU/GPU coolers would allow.
2.) Thermal conductivity of water(other fluids can also be used) is higher than air, allowing more heat to be transfered to a different location more quickly allowing higher overclocks.

That's really all the advantages that I can think of.

 

[clarkjd]

Look at it this way, your PC will give off x number of watts as heat, no matter the method of heat transfer.

 

[4745454b]

I think there are a few more things to consider. Based on that first law, the room would get hotter because you have the added heat from the pump.

I think you also need to take heat capacity into account.

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

If you look at the table water has one of the highest abilities to store heat. When you use a heatsink, you pretty much immediately heat the air and move it to the air outside the case. This will raise the temp of the room quickly. When you use water, you have to use a lot more energy to raise the temp. This energy will not immediately leave the water and join the air. This means that it will take longer to heat the room with water, but if the ambient temp stays constant they should both achieve the same temp.

What makes you ask the question? I get the feeling some background info might help answer whatever it is your looking for.

 

[clarkjd]

I think there are a few more things to consider. Based on that first law, the room would get hotter because you have the added heat from the pump.

I think you also need to take heat capacity into account.

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

If you look at the table water has one of the highest abilities to store heat. When you use a heatsink, you pretty much immediately heat the air and move it to the air outside the case. This will raise the temp of the room quickly. When you use water, you have to use a lot more energy to raise the temp. This energy will not immediately leave the water and join the air. This means that it will take longer to heat the room with water, but if the ambient temp stays constant they should both achieve the same temp.

This is true, but at some point, the heat capacity of water will reach equilibrium, and a watt in will require a watt out, or your CPU temp will rise.

 

[4745454b]

Correct. But it will take time. And by the time it does it will be later in the day/night then the air cooler. Later in the day/night means ambient temps will be lower so the room would be cooler. The flip side is also true. Even with the system off the water will continue to emit heat, meaning morning temps could be warmer. Overall I doubt all this matters much. Just opening the window makes this an open system and changes everything.

 

[amsga]

The reason why I am asking is this is that if your cooling system is more efficient, the core temperature would be lower and hence the processing units need not work as hard. This would then again translate to a not so high core temperature and the cycle some what perpetuates itself.

I understand that there will be a need to achieve thermal equilibrium and I'm not saying that this cycle would result in freezing temperatures or anything like that.

 

[4745454b]

if your cooling system is more efficient, the core temperature would be lower and hence the processing units need not work as hard.

I disagree with this statement. The efficiency of the cooler does not equal lower temps. I could imagine you could have a "90% efficient cooler, but one thats only capable of cooling 100W. Any 100W+ CPU would get to hot whereas an 80% efficient cooler that could handle 200W would be able to keep the temps down. We might be thinking of two different concepts of the word efficiency however.

The other issue is that the temp of the "processing units" doesn't matter to much. As long as the temps are in its operating range, they should work the same. If the CPU is designed to run at -40c to 70c, they should work equally as "hard" at 35c as they do at 60c.

I suggest you don't worry to much about this unless you seriously want to learn. (if so, start doing a lot of reading. Not only on CPUs, but physics, thermal dynamics, and chemistry.) A simpler answer I gave someone recently about coolers is air coolers < heat pipe coolers ~ sealed water coolers < water loop. Lets ignore TEC coolers. These are your options for cooling a CPU for 24/7 use. Pick the one that fits your budget and risk.

 

[jsc]

I think that most of us are using heat pipe type air coolers even of OC rigs that run 24/7.

 

[rubix_1011]

Your CPU/GPUs are generating the same amount of heat watts regardless of the cooling method. Watercooling simply has the capacity to absorb and move more of that heat at any single time than air which is why components stay cooler under water. Radiators also have the ability to dissipate more heat at any single point in time than most air coolers, so it doesn't appear to be as much heat being removed from the system at any given time.

 

[Supermuncher85]

The reason why I am asking is this is that if your cooling system is more efficient, the core temperature would be lower and hence the processing units need not work as hard. This would then again translate to a not so high core temperature and the cycle some what perpetuates itself.

I understand that there will be a need to achieve thermal equilibrium and I'm not saying that this cycle would result in freezing temperatures or anything like that.

I think your missing the point. Heat is an unwanted by product. Although my understanding of CPUs is limited I'll try and simplify things. To calculate data, you have to send information from one point to another. This information is transmitted by electrons. As these move through a material it creates friction which in turn creates heat. The higher the clock, the more electrons need to be sent, the more heat will be created. The type of cooling is irrelevant to the heat given off by CPU. It will not "work" any less because the temperature is lower. The amount of electrons needed to go from point A to point B still remains the same.

Heat will inadvertently cause hardware failure at certain points which is due to Cpu's consisting of many different materials, each having their different thermal properties. All of these parts are interlinked and when one of these parts fails, it usually means the entire CPU fails.Cooling can to some degree lower this risk but there is essentially a maximum electrical resistances which would just instantly vaporize the conducting material. This is also one of the reason why we don't 10Ghz CPU's and instead a transition to multiple cores has taken place. Until there's advances in material that allow for the circumvention of electrical resistance, the maximum performance of CPU's are limited.

If that fails, there's hope quantum computing will come this life time.

 

[amsga]

Best answer selected by amsga.

 

[PeersC]

I am working on a heat exchanger system that should be far more effiecient but it will take time to get perfect - will work from mother nature, store cool thermal energy for whne you need it - will post video on build when ready

Peers