Watercooling CPU temp performance.

Associate
Joined
2 Dec 2012
Posts
134
A while ago after completing my loop I posted that temps were still a little high, since then I have sorted out my CPU block and resat it and now need to work out if something is wrong or I was expecting a little too much.
I have a HWlabs SR-1 360 radiator cooling a i7 [email protected] with a Vcore of 1.28v.

Under prime I will level out at around 68-72c with an idle of 33-40c. While this is not the end of the world it is only around 10-15c less than my air cooler. (to be fair it was a good air cooler). I have 6 fans on push pull and fan speed makes no real difference but it is a low FPI rad.

My thoughts now are

1. I still did not seat the CPU block right but it seems very unlikely now (also used a new thermal compound.

2. I have a poor (in terms of heat) CPU and it is currently more limited by heat transfer rather than the cooler.

3. The low fpi rad is just not cutting it, (seems very unlikely because it does not tend to get all that warm)

I am starting to lean towards option 2 and think I overestimated what water cooling could do (looks very nice however). As a side not one core (0 out of 0-3) is colder than the others. At idle it is around the same but under load can be as much as 11 less. This was the same before and after resitting just an oddity of my CPU I feel.
 
Whats the rest of your config?

Do you only have the CPU in the loop?(no GPUs etc?)
Have you removed the protective cover in the bottom of the CPU block(happens lot of times)
Have you connected the IN and OUT correctly in the CPU block?

It is normal for some CPUs to have some cores higher/lower than others(but 11C is bit much I would say)
 
My setup is very different to yours (an XSPC EX240 with a pair of 500RPM Scythe Kazes and a pair of Zalman Reserators in series cooling a 4GHz i7 3820 and a HD7950 Boost). However I would expect yours to be superior with Just the CPU in the loop and SR-1 360 in push pull. I get sub 30c idle temps, sub 60c prime temps, and sub 50c gaming temps with the GPU flat out.
 
Last edited:
It is only the CPU in the loop, there is no protective cover on the block (I had it apart yesterday to redo it would not have been able to miss it a second time because it took a lot of cleaning. I got the in outs wrong but fixed that yesterday, no real difference.
 
Well, during it's leak it was rocked around and leaned until you could not hear air when the pump was started up. On top of that it got a good rocking when I carried it upstairs. My only benchmark is last time before I re did the loop it did have air trapped (I did not know) so temps at worse got to 78-80c but one morning dropped to about where they are now. So As far as i know, there is not air trapped. I have coloured pipes so cannot be 100% sure.
 
Since you have high temps for a loop with just CPu block on a 360 RAD and 11C difference in temps between cores, I would say time to redo the loop.
 
Any advice on what to do? Yesterday I had the whole loop apart, drained and refilled the fluid, leak tested and cleaned and resat the heat sink, cant think what else to do. I would try another i7 but I can't think where I could get one. All i know as a reference point is that with the dark rock pro 2 and the same settings I would idle around 44-46 and max out around 82c
 
I am thinking as the rad does not get warm either the cpu is not getting heat into water or the water is not getting the heat into the rad. If it's the cpu I am running out of ideas as I have already resat it. As for the rad, I was thinking it is at the top of my case, I could hold the case upside-down for a few seconds that way any air would be sure to get moved out, not sure if this would mess up the pump or other components in any way.
 
how quick do your temps return to idle? my overclocked i5 @ 4.5 would hit 76 degree on 10 run IBT normal, but almost instantly return to near idle temps (usually 24-28 degrees).

When you removed the block was the thermal paste covering the cpu equivalent footprint on the water block? to show you had equal all round pressure and good thermal paste distribution?

no lumps on either cpu or block?

not sure what else to suggest.
 
Temp go back to idle extremely quickly a few seconds at most. The paste looked fine, no different to when I normally do it on air coolers. This is the problem I am limited into what it can be, I am thinking either air in the rad or something wrong with my OC settings that cause a lot of heat output.
 
You also have to remember that the thermal paste used by intel is a bit shoddy with ivybridge (the bit under the metal lid on the processor).

Ask in the cpu section about volts and compare to others, lots have the 3770.
 
Yeah, I will ask the overclocking guys, was torn between that forum or this as it contains a bit of both. There is an option I know that I have a batch that does get quite warm it happens if that's the case no amount of cooling will make much difference.
 
With my bios set to default it sets the processor to 3.9ghz which is more than the processors default but is what the bios wants to be. It sets it to around 1.116vcore (auto but cpu-z got to that under prime) and at idle is much the same as when at 4.6ghz around 30-40. However under load does not go above 54, normally around 52.
 
30-40 idle sounds high, im getting that now with MSN and WoW in the background, but idle isn't that important. If its behaving fine at stock clocks it could just be that your OC has reached the limits of what that chip can do and the Ivy Bridge IHS issue is now coming into play? /shrug.

From what I recall of the issue it gets to a point where the heat cannot transfer fast enough from die to IHS to waterblock and from the description (rad not getting that hot) it sounds like that may be the case as flow wont be the issue with one block and one high flow rad surely.
 
I am leaning towards it being the limitations of the chip at the moment. My overclock does have a high vcore compared to some but I cannot get it stable at any less. My last thought is still a big airlock in the rad as it is the only other way I could explain it, without accepting it's just the chip but I am still not sure about the safety of tipping my PC upside down.
 
Back
Top Bottom