I5 4690k Core Temp Problem

[godsarmy4vr]

So my i5 4690k cores vary in temperature very drastically, sometimes my first core will be at 80C and my 4th core will be at 65C, not to mention 80c is probably near dangerous for the i5 4690k.
Here's a picture of my core temps: http://imgur.com/a/9raqK
I use a 212 EVO, is 80c at 1.26V 4.6GHz normal with this cooler? I assume it's a bit high for the voltage. I reseated my heatsink, and the issue is still there.

 

[geofelt]

It is normal for individual temperatures to vary.
I think you are ok there.

Also, the danger point is around 100c. where the cpu will slow down or even shut itself off to prevent damage.
You are ok there too.

How high you can oc is determined by the quality of your chip.
I think you got an average chip.

You can probably oc higher a notch or two if you can keep the vcore at 1.3v.

That might require a better cooler.
Any cooler needs fresh intake air to work.
Take your case covers off and see if that makes a difference. If it does, look into better case cooling.

 

[godsarmy4vr]

It is normal for individual temperatures to vary.
I think you are ok there.

Also, the danger point is around 100c. where the cpu will slow down or even shut itself off to prevent damage.
You are ok there too.

How high you can oc is determined by the quality of your chip.
I think you got an average chip.

You can probably oc higher a notch or two if you can keep the vcore at 1.3v.

That might require a better cooler.
Any cooler needs fresh intake air to work.
Take your case covers off and see if that makes a difference. If it does, look into better case cooling.

Do you have a source that says 15C in core temp difference is ok? and I don't use a case, its an open air pc.

 

[geofelt]

Google is your friend.
Here is an interesting document I found.
(does not explicitly answer the difference question though)
http://www.tomshardware.com/forum/id-1800828/intel-temperature-guide.html
One explanation for lower core 4 temperatures is that it is the only one that is not sandwiched between other components on the cpu die.

 

[anthonyc821]

So my i5 4690k cores vary in temperature very drastically, sometimes my first core will be at 80C and my 4th core will be at 65C, not to mention 80c is probably near dangerous for the i5 4690k.
Here's a picture of my core temps: http://imgur.com/a/9raqK
I use a 212 EVO, is 80c at 1.26V 4.6GHz normal with this cooler? I assume it's a bit high for the voltage. I reseated my heatsink, and the issue is still there.

Godsarmy4vr,

It can possible be because your thermal paste had spread unevenly and has air pockets, leaving some hot heat in between the cpu and the cooler. Try taking your cpu cooler out, getting alcohol and spraying it onto paper towels and removing the thermal paste. Grab your thermal paste after you remove it from your cpu cooler and cpu and put a small dot in the center and put your cpu cooler. It should spread evenly.

 

[synphul]

If the cpu is consistently running at 80c or higher a better cooler might not be a bad idea. Typically a 212 evo is fine for mild overclocking but 4.6ghz is more at the upper end and more of a scenario for a more adequate cooler. You can try tightening the center knurled screw under the heatsink fins to try and increase contact pressure a bit to see if that helps. Otherwise despite being a good cooler for the money/design, it's still just a 4 heat pipe single fan cooler with a rating to handle up to 180w tdp.

My cpu runs a bit cooler at 4.6ghz with slightly higher vcore (1.284v) but it's being cooled by a dark rock pro 3. It has 7 heatpipes, 2 fans and is rated to 250w tdp. It tops out around 80-81c under p95 stress testing and in the mid/high 60's in games. 99% of the time even under full load it's under 70-75c.

I've got another 4690k with a mild overclock at 4.2ghz running lower temps, lower vcore with a 212 evo. That was before the cryorig h7 was released, had it been available I would have opted for that cooler.

It's also hard to get definitive temps from a stress test like real bench. It's good for a multicomponent full system stress test but due to the varied nature of all the tests the temps are erratic and all over the place. Try prime95 v26.6 with small fft's enabled for a steady state load to get an idea of your thermals. Ideally there wouldn't be more than a 10c spread between any two cores but there's no hard or fast rule or guarantee that there won't be more.

As it is the temps are showing a variation of only 12c overall and the times where you saw 80c on one core and only 65c on the fourth could have been during one of those moments where core one (0) was being stressed harder than the 4th core. That's typical of a normal real working load but why it's important to use something that places even steady load like p95.

http://www.tomshardware.com/forum/id-1800828/intel-temperature-guide.html

 

[godsarmy4vr]

If the cpu is consistently running at 80c or higher a better cooler might not be a bad idea. Typically a 212 evo is fine for mild overclocking but 4.6ghz is more at the upper end and more of a scenario for a more adequate cooler. You can try tightening the center knurled screw under the heatsink fins to try and increase contact pressure a bit to see if that helps. Otherwise despite being a good cooler for the money/design, it's still just a 4 heat pipe single fan cooler with a rating to handle up to 180w tdp.

My cpu runs a bit cooler at 4.6ghz with slightly higher vcore (1.284v) but it's being cooled by a dark rock pro 3. It has 7 heatpipes, 2 fans and is rated to 250w tdp. It tops out around 80-81c under p95 stress testing and in the mid/high 60's in games. 99% of the time even under full load it's under 70-75c.

I've got another 4690k with a mild overclock at 4.2ghz running lower temps, lower vcore with a 212 evo. That was before the cryorig h7 was released, had it been available I would have opted for that cooler.

It's also hard to get definitive temps from a stress test like real bench. It's good for a multicomponent full system stress test but due to the varied nature of all the tests the temps are erratic and all over the place. Try prime95 v26.6 with small fft's enabled for a steady state load to get an idea of your thermals. Ideally there wouldn't be more than a 10c spread between any two cores but there's no hard or fast rule or guarantee that there won't be more.

As it is the temps are showing a variation of only 12c overall and the times where you saw 80c on one core and only 65c on the fourth could have been during one of those moments where core one (0) was being stressed harder than the 4th core. That's typical of a normal real working load but why it's important to use something that places even steady load like p95.

http://www.tomshardware.com/forum/id-1800828/intel-temperature-guide.html

You thoroughly explained the issue at hand which is why I chose you as the best answer, but now I got it to 1.21V@4.6GHz stable and it runs lower then 70c under load, thanks everyone for the help!