5960x + H110i GT seems to be running hotter than it should

[|oR|Spetsnaz]

I have an i7 5960x OC'd to 4.5 Ghz being cooled by a Corsair H110i GT closed loop cooler. I have been having some crashes in BF4 lately and it prompted me to run some cooling tests using RealTemp 3.70 w/ Prime95 v26.6 64 bit using small FFT's. I know I am overclocked, but my CPU seems to be getting quite a bit hotter than it did when I first installed the cooler. The cooler is installed on the top of my Corsair 750D in a 2 fan push configuration. The radiator is clean and I have visually confirmed both fans are working. Can somebody take a look at these screenshots below and tell me if I am taking crazy pills? I know the 5960x can get pretty toasty when overclocked, but I just want to make sure its working within the expected thermal envelope of other users with the same clocks and setups.

Quiet fan / Performance pump profile:

Performance fan & pump profile:

AI Suite Showing OC settings:

Here is a screenshot from when I first installed the cooler. I am fairly certain this was with a 4.5Ghz OC. Not sure about the fan profile:

 

[Lalau2]

I currently run a 5930K using the same cooler on my AsRock X99 OC Formula. Ran the same test as you and my temps are 39 C and under load no higher than 62 C. I would remount the pump assy to the cpu. Make sure your thermal paste is not old. Let the paste cure a couple of days and run the test again. If it is still running hot , i would look into an AIO with a bigger radiator.

c

 

[lee0539]

Also check that your fans are oriented correctly (may be a simple oversight). Is there enough ventilation? 2 X 980 ti will push out a lot of heat into the case so it may be exhausting hot air out which increases temperature.

 

[Ducman69]

I'm skipping ahead here, but 99.9% of the time, its simply a matter of user error:

1) Using expired thermal paste they found in a drawer that was bought in 2003.
2) Using too much thermal paste
3) Applying the thermal paste improperly (X-method or dot-method are ideal), don't use smear method it can create a giant air bubble
4) Forgetting to take off the very thin clear cling plastic off the bottom of the waterblock
5) Mounting the radiator below the block, which can cause air bubbles
6) Fans oriented opposite from intended direction of flow
7) Waterblock improperly seated, remember to tighten in a cross fashion, slowly, and evenly

 

[|oR|Spetsnaz]

I think I'm probably going to back it down to stock speeds and just wait for the 6950x to come out. I figure it won't be too much longer and I can troubleshoot the cooler then.

 

[RazorWind]

What temperatures are you expecting? What I see in the screenshots looks totally fine.

With a 5280K and an H100i with relatively anemic fans, I see mid 30s at idle and maybe around 60 after an hour of continuous intense use.

I would back the overclock off and see if that fixes your crashing issues. It's possible it's not as stable at that speed as you thought it was.

 

[|oR|Spetsnaz]

What temperatures are you expecting? What I see in the screenshots looks totally fine.

With a 5280K and an H100i with relatively anemic fans, I see mid 30s at idle and maybe around 60 after an hour of continuous intense use.

I would back the overclock off and see if that fixes your crashing issues. It's possible it's not as stable at that speed as you thought it was.

Well, it seemed to stay in the 60's range when I first set up the cooler. I ran the ASUS ROG stress test for about 12 hours with no crashes to verify the overclock was stable. I was looking at some overclocking results from Anandtech and it looks like they were seeing around 85c on a Coolermaster Nepton 140XL, which is a single 140mm fan + radiator. I imagine I should be seeing lower temps on a H110i at the same speeds.

 

[RazorWind]

A few things I can think of to check:

1. You're cooling the rad with air from inside the case, right? Are the other case fans spinning?
2. You're not supplying more voltage to the CPU now, are you?
3. Has the ambient temperature in the room where the system lives increased? My house is about 8 to 10 degrees (fahrenheit) warmer in the summer than in the winter, which would account for a lot, but not all, of your difference.

 

[|oR|Spetsnaz]

A few things I can think of to check:

1. You're cooling the rad with air from inside the case, right? Are the other case fans spinning?
2. You're not supplying more voltage to the CPU now, are you?
3. Has the ambient temperature in the room where the system lives increased? My house is about 8 to 10 degrees (fahrenheit) warmer in the summer than in the winter, which would account for a lot, but not all, of your difference.

1. I am cooling with air inside the case. I have 2 x 140mm fans in the front pulling and 1 x 140mm fan in the back exhausting. I have confirmed that they are all working. I do have dual GTX 980 Ti's with ACX coolers, so they are not exhausting the hot air outside the case, but running Prime95 does not touch these cards, so I don't think they are affecting the temps shown from RealTemp in my first post.
2. I have not tweaked the voltage settings at all. I do believe the motherboard is set to automatically adjust voltage because when I am at 4.5 Ghz, my voltage is about 1.3v.
3. The ambient temp is pretty much constantly 75F. It never goes above that.

 

[RazorWind]

Maybe check the torque on the fasteners that hold the CPU jacket down? Reseat the jacket with new TIM, maybe? I found the Corsair mounting arrangement to be a little tricky to get tightened down properly. I would tighten the nuts down and then check them again and they'd be slightly loose.

I'd also make sure that the system is not adding more voltage than I originally configured it to. The system seems to have quite a bit of latitude in terms of how much voltage it actually uses in these new chipsets. It's not like the old days where you'd just set it to something, and that was the setting.

 

[|oR|Spetsnaz]

Maybe check the torque on the fasteners that hold the CPU jacket down? Reseat the jacket with new TIM, maybe? I found the Corsair mounting arrangement to be a little tricky to get tightened down properly. I would tighten the nuts down and then check them again and they'd be slightly loose.

I'd also make sure that the system is not adding more voltage than I originally configured it to. The system seems to have quite a bit of latitude in terms of how much voltage it actually uses in these new chipsets. It's not like the old days where you'd just set it to something, and that was the setting.

I appreciate it. I checked the torque on the fasteners and it is well seated. As far as I can tell, the voltage is never going above ~1.3v.

To be on the safe side, I ordered an H110i GTX from Newegg just now because it appears they are $60 after coupon and possible rebate...$80 without the rebate. Cheap enough that I am willing to give it a go.