Do Crazy Amounts of Radiators Benefit Overclocking?

Toggle sidebar Toggle sidebar
portablesounds

portablesounds

Honorable
Dec 16, 2012
23
0
10,510
So this question has been bugging me for a while now: I know that there's recommended numbers of radiators in a water cooling loop, but if you have, say, a 560mm rad for just your CPU, then does it increase potential overclocking a lot more than a 240mm rad? As far as I can tell, the three basic things that limit an overclock is: amount of watts/volts that can be supplied to the component, the clock speeds of other parts of the computer, and the temperatures. If a MASSIVE water cooling setup is cooling the component, assuming the other two factors are taken care of, is the only limit the fact that there's a theoretical limit where the water cannot carry the amount of heat the component produces? I was considering an insane watercooling build, but I wasn't sure if a thick 4200mm, a thick 280mm, and 2 thick 140mm's would be worth it for the CPU, FSB/CPU voltage regulators, and 2 GPU's (all in a Fractal Design Arc Midi R2 Mid Tower :)). Thanks for any help in advance.
 
Solution
4Ryan6
Regarding radiator water cooling the best you could possibly hope for it is ambient room temperature and it will take a massive radiator cooling field and money just to accomplish that with an overclocked CPU.

You can always go below ambient but you will not achieve that capability of performance water cooling with radiators!

http://www.tomshardware.com/forum/282844-29-peltier-water-cooling

madcratebuilder said: "I think any OC limit is imposed by the chip and board, not temps."
Click to expand...

If that were the case there would be no LN2, or Dry Ice, or Phase Change, cooling record holders would there.

Temperature is definitely a limiting factor when it comes to overclocking, and not only the top end reach, but also the internal CPU...
Sort by date Sort by votes
M

madcratebuilder

Honorable
May 10, 2013
56
0
10,660
You never go below ambient with water, no matter how much rad you have. The more rad the lower your deltaT to a point. If you have deltaT of 10C or less at full load I think any OC limit is imposed by the chip and board, not temps.

 
Upvote 1 Downvote
4Ryan6

4Ryan6

Champion
Apr 14, 2002
20,610
6
52,915
Regarding radiator water cooling the best you could possibly hope for it is ambient room temperature and it will take a massive radiator cooling field and money just to accomplish that with an overclocked CPU.

You can always go below ambient but you will not achieve that capability of performance water cooling with radiators!

http://www.tomshardware.com/forum/282844-29-peltier-water-cooling

madcratebuilder said: "I think any OC limit is imposed by the chip and board, not temps."
Click to expand...

If that were the case there would be no LN2, or Dry Ice, or Phase Change, cooling record holders would there.

Temperature is definitely a limiting factor when it comes to overclocking, and not only the top end reach, but also the internal CPU throttling built into the CPU itself, that you have zero control over, and the majority of overclockers are not even aware of.

portablesounds said: "If a MASSIVE water cooling setup is cooling the component, assuming the other two factors are taken care of, is the only limit the fact that there's a theoretical limit where the water cannot carry the amount of heat the component produces?"
Click to expand...

Yes, If the water temperature itself cannot go below ambient room temperature!

There are cheaper alternatives to running a below ambient water temperature, but it absolutely won't fit in your case plans.

http://www.tomshardware.com/forum/275185-29-exploring-ambient-water-cooling


 
Upvote 1 Downvote
Solution
portablesounds

portablesounds

Honorable
Dec 16, 2012
23
0
10,510
Thank you for the great responses!

Yes, If the water temperature itself cannot go below ambient room temperature!
Click to expand...
I do understand that it is impossible to get sub-ambient with watercooling since the air cooling the rads is room temperature (if I believed that I'd also believe in perpetual motion machines :D). I'm thinking more along the lines of under load the component's individual temperatures don't reach dangerous levels (like 90C). Even if the device is kept at room temperature, I hope to God my room temps never hit 90C :). My point is I assume as you add more and more radiators, just like with overclocking, that extra 100MHz requires more and more power and produces tons more heat to the point where a whole extra 140mm might only have negligible effects. The "theoretical limit" would be where the room temperature water either cannot conduct heat fast enough away from the components or the deltaT hits obscene levels like 50C (which I know is virtually impossible and could cause some serious pressure problems in the loop). I'd love to do phase change cooling if it weren't for the fact that my computer would sound like a refrigerator :).
 
Upvote 0 Downvote
M

madcratebuilder

Honorable
May 10, 2013
56
0
10,660


The OP asked about cooling with rads not LN2 or other devises. If you have adequate rads to start with adding more rad does little to increase the potential OC.

 
Upvote 0 Downvote
portablesounds

portablesounds

Honorable
Dec 16, 2012
23
0
10,510
Yes thank you madcratebuilder. Is there any documentation out there where someone just took a metric butt tonne of rads onto a CPU or GPU and did some objective testing? If there isn't any, I'll just have to try it for myself if I get enough money to do this build :D.
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom