am i missing something or is this guy talking rubbish?

[Reaper 392]

I'm having a discussion on another forum about coolers. this is what he has said so far:

A lot of the time people say their huge tower cooler gets lower temps is because their huge tower cooler is either on a bench case or they haven't run the computer for hours. Air cooled heatsinks build up in temperature over the more time you run it unless your fan is capable of dispelling more heat then your processor is putting into the tower, and that would require fans like kazes, deltas, and vantecs, which are incredibly loud.

Admittedly i havent tested this myself, but surely the same can be said for water coolers? its a simple case of you're putting (heat) energy in, its chucking that energy to the surroundings. It doesnt mater if the transfer method of processor to air is big slab of metal or some water and a small slab of metal mounted elsewhere. it just depends on how much heat the cooler can get rid of.

From reviews that i've seen, the only all in one water cooler that can outdo tower coolers is the H100. everything else can be beaten by a big tower like the Noctua NH-D14 or the thermalright silver arrow to name a couple

No, because a water block carries the heat down the pipes to the radiator, which means the maximum amount of heat that is inside the coolant is much less then the potential heat of a tower cooler. A tower cooler will get hotter and hotter over time without loud, big fans, but a closed loop system will have much longer temperature stability.

A lot of these reviews are benching their towers open air and most do not do something like a 24-48 hour prime92 run, while they mount the closed loops inside the case, and usually use the stock closed loop fans, which although okay are usually not as good as higher grade fans.

can someone explain who is right in this discussion and why please, because i have zero experience with closed loop water cooling

 

[cheechm]

I would have thought it was more to do with how quickly the water or air heats up. It takes 4 times the amount of energy to change water by 1 celcius than it does to change air by the same amount. On that basis a closed loop water system will have more temperature stability purely because it doesn't heat up so quick. However, this also means when it gets to the radiator that it takes longer to cool down.

Also its not about the fan being able to "dispel" heat rather how quick the fan can pass air through the heatsink.

HTH

 

[Uncle Petey]

A lot of the time people say their huge tower cooler gets lower temps is because their huge tower cooler is either on a bench case or they haven't run the computer for hours. Air cooled heatsinks build up in temperature over the more time you run it unless your fan is capable of dispelling more heat then your processor is putting into the tower, and that would require fans like kazes, deltas, and vantecs, which are incredibly loud.

No. Air coolers only build up temperature over time when case airflow is badly set up.
Fast, loud fans are not necessary. Good airflow is.

Admittedly i havent tested this myself, but surely the same can be said for water coolers? its a simple case of you're putting (heat) energy in, its chucking that energy to the surroundings. It doesnt mater if the transfer method of processor to air is big slab of metal or some water and a small slab of metal mounted elsewhere. it just depends on how much heat the cooler can get rid of.

From reviews that i've seen, the only all in one water cooler that can outdo tower coolers is the H100. everything else can be beaten by a big tower like the Noctua NH-D14 or the thermalright silver arrow to name a couple

It's more about the fact that things like the H100 are usually exhausting directly out of the case and therefore avoiding the heat build up issue.
The H100 only narrowly beats top end aircoolers by way of loud, fast fans. Of course, compromises can be made if you're not going all out for the best temps, but a high end aircooler should offer better temps at noticeably lower noise levels.

No, because a water block carries the heat down the pipes to the radiator, which means the maximum amount of heat that is inside the coolant is much less then the potential heat of a tower cooler. A tower cooler will get hotter and hotter over time without loud, big fans, but a closed loop system will have much longer temperature stability.

A lot of these reviews are benching their towers open air and most do not do something like a 24-48 hour prime92 run, while they mount the closed loops inside the case, and usually use the stock closed loop fans, which although okay are usually not as good as higher grade fans.

Again, noisy, fast fans are not needed. Airflow is. Or as doyll would say: 'Airflow, not airblow'.
Is he trying to say that reviews are biased in favour of aircooling?

I have to lol @ doing 24-48hr Prime runs...

He's not completely wrong, but he's not completely right either.
More likely than not, neither am I

Wait for doyll or Helios to add a little wisdom tomorrow

 

[MaxP779]

Well hes right about the part where he says the H100 is the only AIO to beat big air coolers, the H100 needs 2500 RPM fans to accomplish this feat though, stick the same fans on a big air cooler and the big air will not be beaten by any AIO water cooler.

As for the rest of what he says ive no idea.

 

[setter]

No. Air coolers only build up temperature over time when case airflow is badly set up.
Fast, loud fans are not necessary. Good airflow is.

I can vouch for this, i had the same rig in two different cases. A Lian Li A70F full tower, and my current shinobi xl. Airflow in the LL is vastly superior, (virtually non existent in the xl). Saw a bit of an increase in cpu and indeed overall system temps, simply due to the lack of intake on the xl case.

 

[mmj_uk]

A problem I've found with water is that there is nothing to cool the VRM heatsinks which are a source of heat in high power processors so you tend to get throttling without some sort of fan blowing on them.

I've only found VRM cooling to be an issue on overclocked 3930K and Piledriver though.

 

[Yamahahahahaha]

Two different things being talked about here.
1) Heat capacity (the cooler's ability to store heat)
2) Thermal Conduction (the cooler's ability to move heat from the CPU contact area to the heatsink)

The ideal heatsink has just enough heat capacity to hold on to the heat long enough for it to reach all the extremities of the heatsink.
It's no good if the base of the tower is exceptionally hot but the top is at ambient temperature is it?

The reason watercooling can be better is because you can have a much larger radiator away from the top of the CPU. The corollary of airflow is air moving across the surface area of the heatsink. If the heatsink on top of the CPU has the same surface area as a watercooling radiator, then they will perform equally (all other variables being equal).
The trouble watercooling has these days is designing blocks which most efficiently transfer heat from the CPU into the water without restricting waterflow too significantly. Fluid dynamics are much trickier than air.

What the numpty in quotes #1 and #3 is talking about is a passive system. He is correct about passive system performance, but it is not desirable for a system to have high Heat Capacity at the expense of Thermal Capacity, otherwise you're just holding on to more heat than necessary, rather than expelling it!
Ideally you want to put 1w of heat out of your CPU, and for all that heat to be distributed equally over the fins of the heatsink instantly.

 

[olivier renault]

If you have a closed system, of course it will build up. Put some airflow in, and the temps will reach an equilibrium.

 

[panyan]

A problem I've found with water is that there is nothing to cool the VRM heatsinks which are a source of heat in high power processors so you tend to get throttling without some sort of fan blowing on them.

I've only found VRM cooling to be an issue on overclocked 3930K and Piledriver though.

I thought those motherboard blocks took care of that?

 

[Avenged7Fold]

I thought those motherboard blocks took care of that?

Your right, the heat sinks on motherboards tend to be very effective, especially in high end motherboards where fancy, large heat sinks are one of the selling points (mainly because not they look bad ass). They are however designed to be used with an air cooler on the CPU. The VRM and NB heatsinks tend to surround the CPU because an air cooler will circulate the air around them anyway. If you replace the CPU cooler with a waterblock, depending on your system, you can get very little air movement on the northbridge making those heatsinks less effective. For most people including overclockers, it isnt an issue because the VRM's either don't get hot enough to effect performance or they have ahalf decent air flow due to case fans.

Motherboard water blocks are expensive (but look awesome), a cheap way to get airflow on your motherboard is to put a case fan pulling or pushing air near the VRM's or a fan that points down on your MOBO. Before switching to water cooling the MOBO i used something like this on my RAM and it dropped my VRM and NB chip by a few degree's just because there was a lack of air flow. http://www.custom-build-computers.com/image-files/ram-module-1.jpg

TBH though, I think with any mid to high end MOBO released these days and a voltage which isn't crazy high, you don't need anything more than heat sinks even when just using water cooling on the CPU,