How important is the radiator?

[Aditlojs]

When looking at skinnee's testing the difference between an MCR320 which costs $50 and other "high performance" radiators (which cost much more) seems neglible...

With a 463w load and 1000rpm fans your looking at a difference in your Delta T of .45*...

So that brings me to my questions...

How much is a difference of .45* difference worth to you?

and

How important is the radiator?

The differnce to me seems to be minute... You would see better performance dropping your room temp 1*C than you would by switching out one rad for the other. (assuming your not running a high fin density rad with super slow fans)

http://skinneelabs.com/triple-radiator-comparison-v2/4/

Am I wrong here?

 

[x509]

When looking at skinnee's testing the difference between an MCR320 which costs $50 and other "high performance" radiators (which cost much more) seems neglible...

With a 463w load and 1000rpm fans your looking at a difference in your Delta T of .45*...

So that brings me to my questions...

How much is a difference of .45* difference worth to you?

and

How important is the radiator?

The differnce to me seems to be minute... You would see better performance dropping your room temp 1*C than you would by switching out one rad for the other. (assuming your not running a high fin density rad with super slow fans)

http://skinneelabs.com/triple-radiator-comparison-v2/4/

Am I wrong here?

Or how about this? For about the same cost, you could do two Swiftech radiators, or one of the others. At the end of the day, for the same $$$ (or euros), you would get much more cooling, I would think.

 

[Aditlojs]

Or how about this? For about the same cost, you could do two Swiftech radiators, or one of the others. At the end of the day, for the same $$$ (or euros), you would get much more cooling, I would think.

Well then you run into space constraints if you are not using a bench system... Like me!

 

[Cruiza]

No you are not wrong. The premium you are paying is mostly for subjective aesthetics and perhaps some quality differences. The Swiftech rads are about the best bang-for-the-buck as you can get.

 

[dnottis]

I went from a Swiftech 2x120 to a PA120.2 to PA120.3 cooling both CPU and GPU (both overclocked). I could use lower speed fans (which I didnt I use the same medium speed yate) with the thermochill rad and temps were only a few degrees difference.

Knowing what I know now... a dual swiftech rad would be used. The PA models are just so big and take a lot more liquid to fill... making them much heavier as well. When it comes down to it, once you go water you gain the cooling benefits from it..the rest of the hardware is only going to give you a few degrees here or there. From a $45 dual rad to a $120 triple rad I dropped 3-4 degrees at most.

I've even got a few guys at work running Corsair H60s now and with 2 x 120 fans mounted on it the rads their temps are within 7-8 degrees of mine. Considering they paid $58 after MIR and I've got about $350 into water cooled equipment (granted I've got my GPU water cooled too though)...go water, but you don't need to top of the line, most expensive stuff. Slightly more flow and 1 or 2 degrees isn't worth another $80 now that I've been on both sides of it.


I recently fell for it again and moved from my D-Tek Fuzion with home made mouting hardware on 1155 to XSPC Rasa.. and in some reviews I saw 4-5 deg difference... guess what I saw... nothing. lol same temps but I've now got an updated block that is actually compatible with new sockets. It was time to retire the Fuzion but don't fall for the marketing... buy what you can afford.

 

[ghost6303]

on that page you linked to i see measurments of; air temperature, water temperature, and fan RPM.

to measure the efficiency of a radiator you need to measure; water temperature, air temperature, water flow rate, air flow rate.
those are the four main variables which directly effect the efficiency of the radiator. any test that omits any of these four variables cannot be considered complete.


take a look at this picture....
http://skinneelabs.com/assets/images/Radiators/TripleV2/triples-v2_OC920-275sli_deltaT.jpg

lets just look at the first radiator, the HWL GTX 360.

let the D12SL12 have a flow rate of [X] CFM;
the D12SH12 have a flow rate of [Y] CFM;
and the DFS123812H have a flow rate of [Z] CFM,

if x=10 CFM
y=20CFM
z=30CFM

the results would mean something completely different, than if...

x=10CFM
y=50CFM
z=100CFM

but we have no idea because it wasnt measured. air speed does NOT relate directly to fan speed, and fan speed does NOT directly relate to radiator efficiency.

 

[dnottis]

well keep in mind the flow of the liquid will be affected by tubing, how you tube it (using y, 45 deg elbow, etc), and the restrictiveness of the blocks used.

Air flow of the fans is affected by static pressure of the fans and how easily they are able to move the air through the case you are using.

Ambient temp also is a big factor since you will never be able to bring the temp of the loop to lower than ambient.

So if you are actually trying to calculate it there are just too many factors to take into consideration. So while it's nice to see some people actually try to do this, these results may not reflect what everyone else will see.

In my experience and buying decisions based off of others testing like this I've never seen a gain worth spending the extra cash for.

 

[Aditlojs]

I realize there are always different variables at play but skinnee does a pretty good job of eliminating a lot of them...

Taking the same loop and the same fans from the mcr320 to the rx360 or 120.3 yields less than 1*C difference in water temp at 463w of heatload. This is what I see by the last graph...