Is there a real advantage of water cooling over air?

If the xp90c has a C/W of 0.11, according to:

http://www.overclockers.com/articles373/p4sum.asp

and the Swiftech Storm Waterblock has a C/W of 0.12 according to:

http://www.overclockers.com/articles373/wbsum.asp

Since air actually outperforms water...what is the advantage of water then? It can't be sound because the xp90c still performs 0.13 c/w with a quiet fan which beats 99% of waterblocks. It can't be cost because water is at least 3x more expensive than an xp90c.

I thought water's big advantage was superior cooling? I'm confused.

1- higher heat dissipation capacity
2- can be quieter than air
3- can cool multiple components in 1 loop (cpu/gpu/nb, etc)
4- fun (for some people)

Maybe the actual design is better in terms of material mass. But just because an object can absorb and retain more heat, if it does not expel it... it is useless. When you think about the thickness of baseplates of Heatsinks and waterblock, the HSF is always thicker and therefore can retain more heat. But waterblocks take advantage of WATER instead of air to carry that heat away from the source at a more efficient rate. When you get the heat ratings between AIR and WATER, you will see why the waterblocks win out.

Just my thoughts,
Bryan D.

bryan_d said:

Maybe the actual design is better in terms of material mass. But just because an object can absorb and retain more heat, if it does not expel it... it is useless. When you think about the thickness of baseplates of Heatsinks and waterblock, the HSF is always thicker and therefore can retain more heat. But waterblocks take advantage of WATER instead of air to carry that heat away from the source at a more efficient rate. When you get the heat ratings between AIR and WATER, you will see why the waterblocks win out.

Just my thoughts,
Bryan D.

Click to expand...

the air the heatsink is using is the air already heated inside your case, unless ducted very well of course. Thus it is recycling heat in a way. Water cooling gets fresh air from outside the case to cool it. so the temps are going to better than air.

The waterblock C/Ws cannot be directly related to the aircooling C/Ws - the aircooling test bed is a P4 with barely any temp probe, the waterblock one is an actual heat die. In short, dont compare them, because they are non-comparable.

Confused me, too - someone REALLY should change that.

edit

Theres this article
http://www.overclockers.com/articles1259/

which is, at best, a ballpark estimate of what the c/w relation is. Obviously, high end water is significantly better than high end air.

Etacovda said:

Obviously, high end water is significantly better than high end air.

Click to expand...

Thanks, that clears things up a lot.

your comparing tangelos and lemons.

the premis of watercooling is to get the heat way from the cpu as effciently as possible.

when you use a HS the conduction material is not moving the heat away from the cpu very much or efficiently since it relies on the material store and transfer the heat on a relitively small surface area.

water is capable of holding more of a load than air already and transfer between solid and liquide is more efficient than solid to gas because conduction> convection
so you may be led to the " well the core/rad transfers the heat to air" argument but you have several things to consider.

heater cores make use of a LOT more surface area both internally and externally

spreading the heat over a larger area means a less dense heat load ( aka larger area for same heat) which in turn means fans have a easyer time removing the heat at lower speeds . do not confuse this with the DT . they are related but as you know most people dont want to go deaf from using a HSF in the 45db range so larger slower adnd less audible fans can move as much heat because they have more surface area to utilize

when you think about it, the transfer rates that you listed should be fairly close to each other. You are talking about the same material, or element, copper in this matter. The variation is expected with the numerous alloys available.

The benefit of water is the increased surface area that it provides by attaching a larger heat exchanger versus a HSF.

thickness determines transfer rate as much as surface area on the block side of the equation.

thorilan said:

thickness determines transfer rate as much as surface area on the block side of the equation.

Click to expand...

thickness will determine the heat transfer rate through the whole system. The coeffiecient of thermal conductivity for copper will be the same for both AIR and WATER blocks if we are talking about the same alloy. This principle would also apply to the radiator or the HSF.


Thermal conductivity is a constant as far as the material is concerned

Area = total surface area available between the 2 different temperature zones

Length = thickness of material between the 2 temperature zones

Delta T = temperature differential between the 2 zones.


HeatTransfer=(Thermal Conductivity) x(AREA/Length) x(Delta T)

Now we can see that the length or thickness of the material will have an inverse effect on the heat transfer rate.

If I might add, water is also about eight times as conductive as air (temperature wise). It's somewhere around eight, correct me if I'm wrong.

-GearShift3r

I thought I read it was several hundred if not more times thermally conductive than air, but I read that a while ago (and I don't remember exactly where either).

noxqzs you completely missed the point of that last post.

was stressing importance of thickness

After reading your comment, I noticed my initial post was a little vague and misleading.

noxqzs said:

Now we can see that the length or thickness of the material will have an inverse effect on the heat transfer rate.

Click to expand...

In the second post, I am agreeing with you that thickness will also be a big factor in determining heat transfer.