Whats going on with my I7 6700k?

[Aladdad1]

Alright, i got a new system built yesterday i have a 6700k + Asus z170 Gaming aura+ H80i... system is up and running. but my Temperatures are crazy.. my idle temps are 20-25c which is alright.. 

but all of a sudden while browsing chrome they would spike up to 55-60 for a second and back down to 25c..
At first i didnt think much of it but then i started playing overwatch i noticed my temps were fluctuating between 45-75c 

So i try to run a more CPU demanding game like BF1 and The division and i notice the temps to act similarly but between 45c and 80c! at some point it spiked from 55c to 85c for 1 second... 

Temperatures are never consistent a second its 45c another its 70c another its 55c another its 80c.. could it be false reading? i tried using real temp,HWinfo , Corsair link

I want to Overclock this CPU but obviously its not doable like this.
I tried reapplying thermal paste over 5 times with same results.. now could it be my chip? 

*i was using this H80i with an 4690k just 3 days ago and overclocked to 4.4Ghz and never reached over 70c

 

Thanks for your help!

[SteinerW]

try setting the voltage in the motherboard, Some motherboards will give the chips way too much voltage. People have reported up to like 1.4 volts. set it to like 1.3 and see if anything changes. stock voltage is sort of a range rather than a set value.

[Guest]

I would assume it's the mount of the pump/cooler, but it looks like you have eliminated those as variables in the equation. 

 

Has it always ran hot? 

 

Could be bad IHS.. although unlikely. 

Not the best cooler at the end of the day, but I would expect a smedge better temps. 

[Aladdad1]

7 minutes ago, Alexokan said:

I would assume it's the mount of the pump/cooler, but it looks like you have eliminated those as variables in the equation. 

 

Has it always ran hot? 

 

Could be bad TIM.. although unlikely. 

Not the best cooler at the end of the day, but I would expect a smedge better temps. 

Could it be bad TIM? ive had this arctic silver for like 7 moths sitting in the drawer

[SteinerW]

4 minutes ago, Aladdad1 said:

Could it be bad TIM? ive had this arctic silver for like 7 moths sitting in the drawer

unlikely. that stuff sits in warehouses for months.

[Guest]

6 minutes ago, Aladdad1 said:

Could it be bad TIM? ive had this arctic silver for like 7 moths sitting in the drawer

Sorry, I meant IHS 

[Aladdad1]

5 minutes ago, Alexokan said:

Sorry, I meant IHS 

should i RMA the cpu then? 

[Guest]

1 minute ago, Aladdad1 said:

should i RMA the cpu then? 

I doubt that's the problem, but it could be.

 

What's the voltage and clock at (stock clock still?) 

[Aladdad1]

9 minutes ago, Alexokan said:

I doubt that's the problem, but it could be.

 

What's the voltage and clock at (stock clock still?) 

Clock speeds at stock is 4-4.2Ghz i noticed voltage reached up to 1.370v at some point so i went and manually set them to 1.255v(adaptive) "CPUz max read 1.278v" and its been stable since then.. no overclocking yet.. just completed a match in overwatch temps actually spiked to 79/83/75/78

 

updated my bios to latest version, did not help. at this point i can't really think of anything other than its the CPU it self that has problems.

[Guest]

1 minute ago, Aladdad1 said:

Clock speeds at stock is 4-4.2Ghz i noticed voltage reached up to 1.370v at some point so i went and manually set them to 1.255v(adapavtive) "CPUz max read 1.278v" and its been stable since then.. no overclocking yet.. just completed a match in overwatch temps actually spiked to 79/83/75/78

 

updated my bios to latest version, did not help. at this point i can't really think of anything other than its the CPU it self that has problems.

I'd RMA if you think your temps are fucked. Not sure where you got it from. 

 

Definitely high for that voltage and that cooler (don't get carried away thinking that cooler should keep it at sub 60c, however. You are still 10-15% warmer than you should be. 

[pyrojoe34]

1. First thing to check is your mount. The H80i has very stiff tubing which makes it very hard to mount the block flush unless it is entirely out of the system. Take the cooler out of the case, then mount the block with the tubes relaxed so you are positive you get flat, even pressure on the block/CPU interface. Then put it all back into the case. This is a common problem with the Corsair AIOs since the tubes that are very stiff and many people have a hard time mounting them flush (made a 20C difference to me and did the same to plenty of other people I told to try this).

 

2. Is the pump running? If you're using Corsair Link to monitor and control the AIO check and make sure the pump RPM is not 0, you should also be able to feel a mild vibration in the block.

 

3. Are the AIO fans doing anything weird like fluctuating randomly or not going the speed you set them at? If yes, read my signature for the fix to that issue.

[Guest]

3 minutes ago, pyrojoe34 said:

1. First thing to check is your mount. The H80i has very stiff tubing which makes it very hard to mount the block flush unless it is entirely out of the system. Take the cooler out of the case, then mount the block with the tubes relaxed so you are positive you get flat, even pressure on the block/CPU interface. Then put it all back into the case. This is a common problem with the Corsair AIOs since the tubes that are very stiff and many people have a hard time mounting them flush (made a 20C difference to me and did the same to plenty of other people I told to try this).

 

2. Is the pump running? If you're using Corsair Link to monitor and control the AIO check and make sure the pump RPM is not 0, you should also be able to feel a mild vibration in the block.

 

3. Are the AIO fans doing anything weird like fluctuating randomly or not going the speed you set them at? If yes, read my signature for the fix to that issue.

He was just running it and it was performing fine.

 

Tubing could be an issue, or simply not getting power to the pump, etc. But if those things check out, there is something more wrong. 

[pyrojoe34]

Just now, Alexokan said:

He was just running it and it was performing fine.

 

Tubing could be an issue, or simply not getting power to the pump, etc. But if those things check out, there is something more wrong. 

In the old system... he moved it to a new one so any number of things could have changed. A bad mount is still the most likely. Or it could be something like a short, getting knocked, etc. I've had an AIO die by moving it from one build to another and had to RMA it. I didn't do anything to it but the movement could have triggered a defect (bad solder/connection) that was dormant.

[Guest]

1 minute ago, pyrojoe34 said:

In the old system... he moved it to a new one so any number of things could have changed. A bad mount is still the most likely. Or it could be something like a short, getting knocked, etc. I've had an AIO die by moving it from one build to another and had to RMA it. I didn't do anything to it but the movement could have triggered a defect (bad solder/connection) that was dormant.

He says he remounted it 5 times.. I doubt it's the mount. He is sure to get it right, one of those times. 

 

 

[Aladdad1]

I  rested the AIO as i was mounting it yesterday to make sure the system was running before securing the rad in place as i saw the temp in the bios at 23c idle i thought it was good.  the pump is running at around 2300rpm and the fans are working and being detected by corsair link. i tried multipile times after reapplying thermal paste to leave the tubes straightened to insure good fluid flow still same temp results.

 

just for a test i tried opening 2 twitch streams and overwatch at the same time after a match my temp spiked to 88c on one of the cores, i don't think that is normal at all is it? 

 

*edit* as i mentioned originally i was using this cooler on an 4690k that was overclocked to 4.4Ghz @1.260v  and it managed to never go over 70c under stress test load while on stock the temps used to top out at 58c... 

[pyrojoe34]

2 minutes ago, Aladdad1 said:

I  rested the AIO as i was mounting it yesterday to make sure the system was running before securing the rad in place as i saw the temp in the bios at 23c idle i thought it was good.  the pump is running at around 2300rpm and the fans are working and being detected by corsair link. i tried multipile times after reapplying thermal paste to leave the tubes straightened to insure good fluid flow still same temp results.

 

just for a test i tried opening 2 twitch streams and overwatch at the same time after a match my temp spiked to 88c on one of the cores, i don't think that is normal at all is it? 

Idle doesn't tell much about a bad mount. There is very little heat to pull from a CPU at idle so you wouldn't know if the mount is bad. You can even (temporarily) have no cooler on a CPU on idle (like in the BIOS) and the temps will stay in line. When I had a bad mount my idle temps were <30C but as soon as I ran anything they'd rocket up in a second. Remounting it fixed that issue, the problem was not bent tubes but the fact that I had mounted it while the tubes were not totally relaxed. I remounted it a few times without it working but then took the entire rad out of the case, held the cooler over the board so there was no pressure on the block except for my finger pressing down in the middle while I evenly tightened the screws diagonally (quarter turn at a time so the pressure is even).

 

I'm not saying this is definitely your problem but I've seen it too many times and your symptoms sound exactly like everyone elses.

 

What are your water temps? If the CPU temp goes up disproportionally to the water temp (they should be pretty well correlated, the delta on my CPU to water temp is 20C or less) then the issue is bad heat transfer from the CPU to the block.

[Aladdad1]

19 minutes ago, pyrojoe34 said:

In the old system... he moved it to a new one so any number of things could have changed. A bad mount is still the most likely. Or it could be something like a short, getting knocked, etc. I've had an AIO die by moving it from one build to another and had to RMA it. I didn't do anything to it but the movement could have triggered a defect (bad solder/connection) that was dormant.

Unsure where to see the water temp but im assuming its the H80i temp in corsair link? its showing a max of 54C (after a couple of overwatch matchs).. never hit that high before previously the highest i've seen was around the 45c. Speaking of heat i just completed a couple of overwatch matches temps toppd at 84c so i figured i'll run P95 26.6 and temps instantly hit 91c so i stopped it immediately, i can't leave the test running in order to let the water catch up with that temp because its too dangerous

[Aladdad1]

I'm gonna head to the store where i purchased these parts from (hopefully they cooperate) and see if they can test another cooler or maybe another CPU to check what is the problem exactly 

[pyrojoe34]

17 minutes ago, Aladdad1 said:

Unsure where to see the water temp but im assuming its the H80i temp in corsair link? its showing a max of 54C (after a couple of overwatch matchs).. never hit that high before previously the highest i've seen was around the 45c. Speaking of heat i just completed a couple of overwatch matches temps toppd at 84c so i figured i'll run P95 26.6 and temps instantly hit 91c so i stopped it immediately, i can't leave the test running in order to let the water catch up with that temp because its too dangerous

54C for the water temp is very high. You generally don't want water temps to get above 50C and ideally lower than 45C. However, those high temps tell me you're probably getting decent heat transfer to the cooler (unless the high temps are due to already hot air going through the rad in which case you could still also have bad heat transfer).

 

Water temps that high are usually one of three things. Either the pump isn't running and the hot water is just staying by the sensor rather than cycling through the loop, or the air going through the rad is already hot and won't cool the fins much, or the cooler is overworked and your OC is too high for the cooler to handle. To be fair the h80i is a 120mm cooler so it can't actually dissipate all that much heat compared to a larger AIO. If you want to really OC your CPU you should be looking into at least a 240 or 280mm rad to handle all that heat.

 

Is your cooler intake or exhaust? If it's exhaust you should try flipping the fans to intake. It could also be that your case temp is pretty high and running hot air through the rad will make it hard to cool any water (the bigger the delta between air and water temp the better). If it isn't already intake then make it intake and see what you get from that.

[smokefest]

These temp are not so bad. A bit high for h80 but heh.. Skylake runs hot.

But my coolermaster evo 212 seems to do a better job than ur h80 wich is weird..

FYI I run 6700k OC @ 4.5 ghz @ 1.335 Vcore with coolermaster evo 212 and in BF1 after hours of gameplay max temp was 76c

 

If I was you I would take off heatsink , apply good offmarket thermal paste ( artic silver 5 ) and reseat heatsink, retest temp after that.

[Guest]

advice in this thread is baffling after all the evidence OP has given. 

 

 

[Aladdad1]

45 minutes ago, smokefest said:

These temp are not so bad. A bit high for h80 but heh.. Skylake runs hot.

But my coolermaster evo 212 seems to do a better job than ur h80 wich is weird..

FYI I run 6700k OC @ 4.5 ghz @ 1.335 Vcore with coolermaster evo 212 and in BF1 after hours of gameplay max temp was 76c

 

If I was you I would take off heatsink , apply good offmarket thermal paste ( artic silver 5 ) and reseat heatsink, retest temp after that.

There! I'm pretty sure the H80i can beat a Evo 212 even by 1-5% I play Overwatch and the temps reach 85c after 30 minutes and not even reach but it "spikes up to 85c" meaning it would go there for a second then back to low 60's and at some point down to high 40's and so on.. its not consistent. The performance is great! but the temps are very bad. After an hour of BF1 i saw my temp spiked up to 91c... i let P95 running small FFT test and the temps actually spiked up to 98c!  throttling since it only spiked there and went back to low 60's Took my PC to the local shop *the one i got the CPU from* he said it may be the CPU it self he will look into it. 

 

NOTE: CPU is on STOCK speeds undervolted to 1.265v as it used to hit around 1.370v on auto

[Aladdad1]

Okay, turn out it was the H80i that seems to have died.. or something happened to it. Even though Corsair link showed normal reading it was not cooling my CPU properly no matter what i did... Tighten the backplate the cpu block re apply thermal paste nothing helped. ended up buying an EVO 212 instead and honestly this thing is a lot better than i expected to be keeping my CPU under 74c during P95 @ 4.5Ghz... Sorry to bother you guys with my trouble, thanks for everyone who tried to help!