Temperature issues with i7-2600k @ stock

[Caimath]

Well, I'm having a bit of an issue right now. I finally decided to enable the XMP profile in my BIOS, after running my PC for over 2 years now. Unfortunately doing this, increased my temps by more than 20°C, idling at around 60°C. After doing some googling, I found that I'm not the only one experiencing this issue and that it is, possibly, caused by the BIOS basically disabling the Intel Speedstep technology causing my CPU to idle at a full 3.8 GHz. After that, I managed to find a guide for manually re-enabling the speedstep technology. Doing this reduced my idle temps back down to around 35°C, so I, of course, assumed everything was alright again.

 

Later I wanted to do some gaming, fired up Borderlands 2 and started shooting things. After about 5-10 mins of running, I started noticing strange framedrops. Thinking it was my GPU OC acting up, I closed down the game and took a look at MSI Afterburner and started scrolling through the hardware monitor graphs. This is where I noticed my CPU temps had maxed out at almost 100° C. Seeing this I immediately shut down the PC, entered the BIOS and restored optimized default settings. However, after this my current problem happened.

 

Basically my temps are completely screwed up, even with default BIOS settings. It idles at around 50-55°C and quickly rises to post 90°C, at any kind of load (opening Chrome with around 20 tabs bumps it up to nearly 70°C ).

 

I'm using the stock cooler, with little issues before this, because I never overclocked the CPU.

 

My question to you is this:

 

Is it possible that the 2 years old stock thermal compound burnt out, due to the extremely high temperatures?

 

 

 

Either way a new cooler is on the way, hoping this can alleviate the issues.

[FireFox]

You could try replacing the thermal paste and clearing out any dust that is in your heatsink and case but thermal paste should last longer than 2 years. What is your RAM voltage and CPU voltage?

[Sauron]

I did the same thing some time ago, you need to go into the bios and enable turbo boost. That should fix it.

[Caimath]

You could try replacing the thermal paste and clearing out any dust that is in your heatsink and case but thermal paste should last longer than 2 years. What is your RAM voltage and CPU voltage?

After restoring to the default settings CPU voltage is 1.240 and RAM voltage is at 1.5 (XMP profile raises this to 1.65). XMP didn't seem to change the CPU voltage, if the voltage monitor in the BIOS is accurate.

 

Unfortunately I don't have any replacement thermal compound around, so I can't really replace it at the moment. I don't think it's caused by excessive dust either, as I cleaned the entire case not too long ago, and all of this happened within an hour or so (I doubt an hour's worth of dust would cause 20°C differences).

[SkilledRebuilds]

 

Either way a new cooler is on the way, hoping this can alleviate the issues.

No need, here is a link to some youtube guides on how to clean your intel heatsink, and re-paste the thermal grease/paste very easily.

https://www.youtube.com/results?search_query=replace+thermal+paste

[Caimath]

No need, here is a link to some youtube guides on how to clean your intel heatsink, and re-paste the thermal grease/paste very easily.

https://www.youtube.com/results?search_query=replace+thermal+paste

I have been wanting to replace the stock cooler for ages anyway, I'm looking to start doing some light overclocking soon(ish). Thanks for the link though.

I did the same thing some time ago, you need to go into the bios and enable turbo boost. That should fix it.

Turbo Boost seems to be enabled already, or at least I think it is (I can only find Turbo Mode, I'm assuming this is the same thing).

I'm a bit of a noob, when it comes to BIOS manipulation, unfortunately.

[Sauron]

Turbo Boost seems to be enabled already, or at least I think it is (I can only find Turbo Mode, I'm assuming this is the same thing).

 

I'm a bit of a noob, when it comes to BIOS manipulation, unfortunately.

 

Try resetting the bios if you don't know what changed. There should be an option for intel speedstepping too.

[Caimath]

Try resetting the bios if you don't know what changed. There should be an option for intel speedstepping too.

If by resetting the bios, you mean restoring to default settings, I have already done that. This dropped the idle temps a tiny bit, but it still reaches near 100°C at load.

 

If you mean something else, I have no idea how to do it.

[FireFox]

If by resetting the bios, you mean restoring to default settings, I have already done that. This dropped the idle temps a tiny bit, but it still reaches near 100°C at load.

 

If you mean something else, I have no idea how to do it.

 

What else have you done to the computer?

 

If you've only been in the BIOS and now that is back to how it was but you're still reaching those high temperatures then is there a possibility that you've done something else inside the case?

[Sauron]

If by resetting the bios, you mean restoring to default settings, I have already done that. This dropped the idle temps a tiny bit, but it still reaches near 100°C at load.

 

If you mean something else, I have no idea how to do it.

 

Try re-applying thermal paste, maybe it got burned off. Check what clock speeds you are getting under load with cpu-z

[Caimath]

What else have you done to the computer?

 

If you've only been in the BIOS and now that is back to how it was but you're still reaching those high temperatures then is there a possibility that you've done something else inside the case?

I haven't done anything else inside the case. Would a dying case fan (I only have mounting slots for 2 120mm fans in my case, side panel intake and rear exhaust) cause increases of more than 20°C? They are both still spinning, but it's impossible for me to check if they are still running at full speed (they are both molex fans, the case is ancient. )

 

 

Try re-applying thermal paste, maybe it got burned off. Check what clock speeds you are getting under load with cpu-z

Unfortunately I don't have any replacement thermal paste around.

 

Load clock speeds are at 3.8 GHz, which is the normal rated turbo clock. It seems the Core Voltage fluctuates a bit though, from between 1.232 to 1.253. Is this normal?

[FireFox]

I haven't done anything else inside the case. Would a dying case fan (I only have mounting slots for 2 120mm fans in my case, side panel intake and rear exhaust) cause increases of more than 20°C? They are both still spinning, but it's impossible for me to check if they are still running at full speed (they are both molex fans, the case is ancient. )

 

Even if both fans died completely they wouldn't cause the temperatures to rise by 20 degrees, well like someone has already said you should replace the thermal paste and see if that has any effect, if it doesn't then I have no idea what could be the issue.

[Caimath]

Even if both fans died completely they wouldn't cause the temperatures to rise by 20 degrees, well like someone has already said you should replace the thermal paste and see if that has any effect, if it doesn't then I have no idea what could be the issue.

Yeah, thats what I've been thinking. I'm gonna replace the stock cooler as soon as the new one arrives anyway. If the issue is caused by the thermal compound, the cooler replacement should fix it (and give me better cooling performance on top of it).

 

Thanks for the help everyone. I really appreciate it.

[Sauron]

Load clock speeds are at 3.8 GHz, which is the normal rated turbo clock. It seems the Core Voltage fluctuates a bit though, from between 1.232 to 1.253. Is this normal?

 

It isn't. You might have damaged the motherboard. The stock voltage should be 1.165. Try manually changing it to that.

[Caimath]

It isn't. You might have damaged the motherboard. The stock voltage should be 1.165. Try manually changing it to that.

My chip was at 1.240v at stock, apparantly it differs from chip to chip. I have tried manually changing it to 1.240, it didn't make any difference on the temps.

[Sauron]

My chip was at 1.240v at stock, apparantly it differs from chip to chip. I have tried manually changing it to 1.240, it didn't make any difference on the temps.

have you tried 1.165?

[Caimath]

have you tried 1.165?

I have not, but that would be undervolting from the stock voltage intel set my chip at. I have no idea what effects this might have on things, maybe my chip needs 1.240 to run stable at stock speeds?

[Sauron]

I have not, but that would be undervolting from the stock voltage intel set my chip at. I have no idea what effects this might have on things, maybe my chip needs 1.240 to run stable at stock speeds?

 

Do give undervolting a try, if it is stable at a lower voltage is should lower your temperatures a fair bit. If it's unstable, it will either revert to previous conditions by itself or resetting the bios will do that for you. This is a temporary solution though, you should try to get access to some thermal paste as soon as possible so you can determine if that's the problem.

[Caimath]

Do give undervolting a try, if it is stable at a lower voltage is should lower your temperatures a fair bit. If it's unstable, it will either revert to previous conditions by itself or resetting the bios will do that for you. This is a temporary solution though, you should try to get access to some thermal paste as soon as possible so you can determine if that's the problem.

Yeah, I will. Thanks for all the suggestions.

[airdeano]

@Caimath

ive had my sandy bridge 2700k on 3.9GHz with 1.095v. on 28° ambient air i had

went from 69° 100% load @ 1.26v to 55° 100% load @ 1.095v so yes, it'll undervolt

just lock your multi to 37 and start lowering the voltage to find the undervolt to run

max turbo speed. due to the silicone placement, some take less voltage to run and

some take normal voltage. mobo mannies overvolt to cover the whole gambit of

silicone used. so AUTO voltage usually means overvoltage.

 

even on Haswell i7-4770k i'm 3.9GHz on .95v 100% load with XTU/Real Bench

[Caimath]

@Caimath

ive had my sandy bridge 2700k on 3.9GHz with 1.095v. on 28° ambient air i had

went from 69° 100% load @ 1.26v to 55° 100% load @ 1.095v so yes, it'll undervolt

just lock your multi to 37 and start lowering the voltage to find the undervolt to run

max turbo speed. due to the silicone placement, some take less voltage to run and

some take normal voltage. mobo mannies overvolt to cover the whole gambit of

silicone used. so AUTO voltage usually means overvoltage.

 

even on Haswell i7-4770k i'm 3.9GHz on .95v 100% load with XTU/Real Bench

I see, this does make sense. I'll give it a try, I'll probably wait for the new cooler / paste to arrive before trying this though. I'm a bit scared of using the PC atm.

[Caimath]

Necroing my own thread just to confirm the solution.

 

It seems that replacing the Thermal Paste did indeed fix my temperature issues.

 

(I know it took me a while to get it done, I've been busy with other things )