4770k@4200mhz 86c w/212 Evo cooling

[jarablue]

I think I may have a bad 212 Evo air cooler. I am running 1.200 voltage and my 4770k is at 4200mhz. I am already seeing under max load of AIDA64 fpu test 86c..highest core got 89c.

Is that normal? I have a CoolerMaster 212 Evo heatsink. I run all the fans in my system at max. The fan on the 212 is running around 1800rpms. I have very good ambient room temps and my idle cpu temps are 24-29c. I have reseated and reapplied AS5 thermal paste till my face turned blue. I made sure I screwed down the heatsink correctly in a cross pattern.

What can I do?

Thanks!

 

[Jaydip]

What is your Prime95 temps?

 

[Arkaign]

Sounds like you lost the TIM lottery. Sorry to say

You could try lapping the 212, but I doubt it will do anything for you. There's also a chance you have a concave or convex IHS, but that's not nearly as common as plain old bad lottery luck with Ivy/Haswell TIM.

 

[JimmiG]

Sounds pretty normal for Haswell. The FPU-only test in Aida64 is one of the worst for temps. Many just ignore that test and run it together with the Aida64 CPU-test to keep temps down.

The clock speed is pretty much irrelevant when discussing temps. The voltage is what matters. 1.2 - 1.25V is about as high as you can go unless you delid. The difference is that some will do 4.6+ GHz at this voltage while others will only do 4.2 GHz.

Before delid I would get about 87C in Aida64 FPU test at 4.3 GHz and 1.218V. Delidding will drop the temp about 20C, allowing you to use voltages around 1.3 - 1.32V.

Switching to MX4 might drop your temp 2-4C also. AS5 is no longer among the top thermal pastes.

 

[CHADBOGA]

I think I may have a bad 212 Evo air cooler. I am running 1.200 voltage and my 4770k is at 4200mhz. I am already seeing under max load of AIDA64 fpu test 86c..highest core got 89c.

Is that normal? I have a CoolerMaster 212 Evo heatsink. I run all the fans in my system at max. The fan on the 212 is running around 1800rpms. I have very good ambient room temps and my idle cpu temps are 24-29c. I have reseated and reapplied AS5 thermal paste till my face turned blue. I made sure I screwed down the heatsink correctly in a cross pattern.

What can I do?

Thanks!

It might not be the solution you are hoping for or want to hear, but what difference in temps do you get if you run it at 4Ghz, instead of 4.2Ghz?

 

[Ed1]

Sounds pretty normal for Haswell. The FPU-only test in Aida64 is one of the worst for temps. Many just ignore that test and run it together with the Aida64 CPU-test to keep temps down.

The clock speed is pretty much irrelevant when discussing temps. The voltage is what matters. 1.2 - 1.25V is about as high as you can go unless you delid. The difference is that some will do 4.6+ GHz at this voltage while others will only do 4.2 GHz.

Before delid I would get about 87C in Aida64 FPU test at 4.3 GHz and 1.218V. Delidding will drop the temp about 20C, allowing you to use voltages around 1.3 - 1.32V.

Switching to MX4 might drop your temp 2-4C also. AS5 is no longer among the top thermal pastes.

Just curious , what cooling did you use to get those results ?
Is that on air or water .

 

[Kenmitch]

I think I may have a bad 212 Evo air cooler. I am running 1.200 voltage and my 4770k is at 4200mhz. I am already seeing under max load of AIDA64 fpu test 86c..highest core got 89c.

Is that normal? I have a CoolerMaster 212 Evo heatsink. I run all the fans in my system at max. The fan on the 212 is running around 1800rpms. I have very good ambient room temps and my idle cpu temps are 24-29c. I have reseated and reapplied AS5 thermal paste till my face turned blue. I made sure I screwed down the heatsink correctly in a cross pattern.

What can I do?

Thanks!

Looks kinda normal to me. Real world temps are what really matter anyways.

Try Linpack 11 if you want to see worse case scenario on temps. Doubt it's possible to ever top Linpack 11 temps short of fan or pump failure.

Linpack 11 generates some crazy temps. But it seems to be good for stabilizing overclocks. Still in the data grabbing/chart building phase but I'm going to generate a new thread very soon. Going to show Linpack 11 vs real world uses for many underclocks/overclocks with some benchmarks and various scenarios. Should give a person an idea of how efficient Haswell is all things considered. Night and day so far.

 

[hillrock100]

It is normal for 4770k at 1.2v

 

[JimmiG]

Just curious , what cooling did you use to get those results ?
Is that on air or water .

Phanteks PH-TC14PE air cooler.

I don't think the cooler matters much if you don't delid. You can see here how close all the coolers are:
http://www.tomshardware.com/reviews/best-heat-sink-haswell,3554-22.html

Even the worst cooler in that test would beat the best one by about 10C if the CPU was delidded.

Before delid, my heatsink would actually be cool to the touch even as the CPU was running 87+C. Now the cooler itself as well as the exhaust air from the rear fan just behind the cooler are actually warm to the touch.

 

[Ed1]

Phanteks PH-TC14PE air cooler.

I don't think the cooler matters much if you don't delid. You can see here how close all the coolers are:
http://www.tomshardware.com/reviews/best-heat-sink-haswell,3554-22.html

Even the worst cooler in that test would beat the best one by about 10C if the CPU was delidded.

Before delid, my heatsink would actually be cool to the touch even as the CPU was running 87+C. Now the cooler itself as well as the exhaust air from the rear fan just behind the cooler are actually warm to the touch.

Yes, I agree if you check many reviews you find 212evo to H100i something like 8-10c higher .
this on 3xxx/4xxx as long as you stay in the 77-85w TDP , sandy-E will be more .
I was just curious as the cooler you had .

 

[JimmiG]

Yep
Also, 89C is nothing to worry about in the Aida FPU test. Fire up a game and monitor your temps. Don't be surprised if they're under 65C.

 

[tweakboy]

I shure hope Ivy E runs cooler then ivy and sandy...