Problem with i7 7700K

[SleVeran]

Hey guys , I recently bought a new computer with the following specs :

Core i7 7700K

Asus ROG STRIX 1070 GTX

MSI Z270 M3 Motherboard

16GB Hyper X Savage CL12

Corsair H100i V2

5 Magnetic Levitation fans (3x 120mm , 2x 140mm)

Corsair 400C Carbide

500GB Samsung 850 EVO SSD

RM750x 80+ Gold

 

It's worth noting at this point that I did not change any of the BIOS settings , CPU voltage is adaptive as default is and Intel boost is enabled by default.

 

The problem is that once i try to run RealBench stress test it makes my CPU usage go up as expected but once it hits really high numbers my PC just freezes for a second or two (unable to move the mouse or anything like that). Plus while i was running those tests and CPU was trying to overclock itself with Intel Boost It got to 4.5Ghz and stayed there pretty solidly , although running RealBench 2.43 made some crashes and applications stopped responding in addition to the occasional freezes.

I checked temperatures and they also seem to be a bit odd , even when just starting up games , before even getting to serious processing power needs it already jumps up to 75C from 40C and the moment i close the game It'd drop back down to  40C in a few seconds??

Also it's worth noting that I have tried running Intel CPU Diagnostics test and my CPU seemed to pass but the same issue happened where it would freeze as i was running the test very often (Multiple freezes throughout the test) Just about the same as it happened running Realbench.

 

So i guess I have two problems :

CPU temps seem to be extremely unstable and change so fast that it looks wrong to me

during CPU intensive benchmarks my PC would freeze completely , only once i tried running RealBench up to a point where my PC stopped responding on some applications.

 

 

whats weird to me is that the CPUTIN sensor (which is apparently the CPU temp sensor from the motherboard shows slow increments every time it refreshes as i start running a benchmark and it locks around 70C after a while without moving , whereas on the very MILLISECOND I press to start a benchmark or a stress test , the CPU package sensor goes INSTANTLY to 90C  (90C is an example of prime95 test)

Here's a video to demonstrate the occurrence : i7 7700K Unstable CPU Temp - YouTube

Watch in the video the CPUTIN Sensor that is the sensor of the CPU provided by the motherboard how it changes temperature in a much more logical way whereas the CPU package sensor goes 40 degress + and then 40 degrees - In 0 time.

I should also mention that because my fans are being controller by the CPU package sensor that means that they go extremely high on RPM when the benchmark is running but as soon as i stop it , it would go to a low RPM which doesn't make sense to me.

Now lets assume that the CPUTIN sensor is a good indicator and the CPU package one is faulty , that means that when my cpu temp rises slowly according to CPUTIN and settles around 70C , it stays there a long time after i finish the stress test just because the fans have already went back to low rpms and barely help cooling it down since the CPU package sensor is lying (at least that's what i think) and claims that the cpu is already cool.

 

BTW this kind of behavior is relevant to every single piece of software that's making my CPU work , not only Stress tests (prime95 or realbench)

 

 

 

I ran the newest version of Prime95 for a few minutes and my cpu was running around 90C but as soon as i clicked stop the test BOOM it went right back into 40C in no time.. how does this make sense..

 

I tried to run Memtest64 I tried to check my SSD using samsung magician , I tried testing my GPU and actually my benchmarks from Fire Strike gave me significantly better performances than other people with the same set of gpu and cpu. BUT i'm still bothered by the fact that it does not seem like i can actually control my fans or monitor the true temp on my CPU... 

 

Edit : 

I just remounted the pump today.. i even used some o rings behind the mount to make the fit tighter on the cpu , btw i just noticed my Vcore moves between 0.928 - 1.288 V and my VID from 0.658 sometimes to 1.345  is that even acceptable?

Thank you to anyone who helps and replies in advance!

 

 

[HairyPotter]

when you first got the pc did you apply thermal paste to the CPU?

[SleVeran]

At first I didn't since the H100i V2 comes with preapplied Thermal compound so I just put it on , but since then i found out the mounting bracket on the H100i V2 has some problem that people complained about so i took care of it by taking it off and putting Spacers on it in order to make the bracket and the water block a tight fit. When i did that fix to the bracket i also applied my own new Thermal compound , of course after cleaning the previous one.

[Cyracus]

P 95 makes temps skyrocket, 90c isn't very surprising with that. as for the temp spiking when you start a stress test and dropping off when you stop it that's normal, it's a rapid onset heavy load and rapid removal of that load so the temps change quickly. The voltage changes are normal as well. But the stuttering may be an issue with tasks running while you stress test fighting for cpu resources or hints of instability, set cpu voltage to auto for a moment and see how it acts. 

[SleVeran]

Wait , if you're saying the literally 1millisecond time from 90C when running a task to 40C when not running a task is normal then is my cooling system even doing anything?

What'd happen if i completely stopped all fans and the pump?

Why do i have the feeling that it really wouldn't make any difference to the behavior of this temperature sensor?

And .. the CPU voltage is set for Auto right now , i did say the Bios are at their default and thats what it's set to.

 

Btw Cyracus have you watched the video i linked ? do you think that behavior is normal?

[faziten]

5 hours ago, SleVeran said:

Wait , if you're saying the literally 1millisecond time from 90C when running a task to 40C when not running a task is normal then is my cooling system even doing anything?

What'd happen if i completely stopped all fans and the pump?

Why do i have the feeling that it really wouldn't make any difference to the behavior of this temperature sensor?

And .. the CPU voltage is set for Auto right now , i did say the Bios are at their default and thats what it's set to.

 

Btw Cyracus have you watched the video i linked ? do you think that behavior is normal?

First and foremost, nobody will recommend you to use Prime 95. It's not even a CPU test anymore, it's more like a VRM ultimate stress, plus an instant way to make your pc crash quick if testing things. By no means is a proper piece of software to test a gaming rig or anything else beyond a VRM burn in. 

 

About temperatures climbing fast: Yes it's normal to see a 30ºC instant jump. Not quite sure about 90ºC i'd blame the adaptive voltage giving in too much so it doesn't crash. Prime95 likes to mess up with voltage so it may give even more. The jump is fast, because your cores are tiny and you are making them push more than 100Amps from one millisecond to the other. So yeah, expect some quick jumps. Package does not grow so fast because it messures the temp on the IHS or the back of the motherboard depending on where it is. Since KrappyLake have dung as Thermal interface material between the core and the IHS, the difference in temperature is to be expected.

Theoretically if the TIM had perfect thermal conductivity and the IHS had zero thermal resistance, those temps would be identical, but it's not the case. Package and core will always have different raw temp readings, even more with Intel mainstream TIM. Expect it to be worse with time, since TIM is not a perfect solution and will degrade through the years, but this is just the tin foil hat. You'll probably sell it before it happens. 

 

Your AIO cooler doing something? Well, it probably is doing something, but you can check. If the pump is warm but the tubes are not, then it's a pump failure (which i doubt, those are normal temps with 7700K). If you stop it, you will see a temperature runaway scenario and your PC will shutdown hopefully before busting something in the process. Check your fluid temp, at the beginning should be cold, that's ok, after few minutes should start to warm up, liquid cooling is not so straight forward as air cooling. the fact that your Temps go back to where they begun when you stop the test, means it's a healthy AIO functioning properly. But hey check it nonetheless. 

 

Why do you have the feeling of blabla... well you are kind of mad? not "insane" i mean angry. It will make a difference try it if you wish. Just don't blame me for anything you do to your rig  

 

 Prime95, 7700K and auto voltage mean HOT temperatures. 

 

 

[SleVeran]

8 minutes ago, faziten said:

First and foremost, nobody will recommend you to use Prime 95. It's not even a CPU test anymore, it's more like a VRM ultimate stress, plus an instant way to make your pc crash quick if testing things. By no means is a proper piece of software to test a gaming rig or anything else beyond a VRM burn in. 

 

About temperatures climbing fast: Yes it's normal to see a 30ºC instant jump. Not quite sure about 90ºC i'd blame the adaptive voltage giving in too much so it doesn't crash. Prime95 likes to mess up with voltage so it may give even more. The jump is fast, because your cores are tiny and you are making them push more than 100Amps from one millisecond to the other. So yeah, expect some quick jumps. Package does not grow so fast because it messures the temp on the IHS or the back of the motherboard depending on where it is. Since KrappyLake have dung as Thermal interface material between the core and the IHS, the difference in temperature is to be expected.

Theoretically if the TIM had perfect thermal conductivity and the IHS had zero thermal resistance, those temps would be identical, but it's not the case. Package and core will always have different raw temp readings, even more with Intel mainstream TIM. Expect it to be worse with time, since TIM is not a perfect solution and will degrade through the years, but this is just the tin foil hat. You'll probably sell it before it happens. 

 

Your AIO cooler doing something? Well, it probably is doing something, but you can check. If the pump is warm but the tubes are not, then it's a pump failure (which i doubt, those are normal temps with 7700K). If you stop it, you will see a temperature runaway scenario and your PC will shutdown hopefully before busting something in the process. Check your fluid temp, at the beginning should be cold, that's ok, after few minutes should start to warm up, liquid cooling is not so straight forward as air cooling. the fact that your Temps go back to where they begun when you stop the test, means it's a healthy AIO functioning properly. But hey check it nonetheless. 

 

Why do you have the feeling of blabla... well you are kind of mad? not "insane" i mean angry. It will make a difference try it if you wish. Just don't blame me for anything you do to your rig  

 

 Prime95, 7700K and auto voltage mean HOT temperatures. 

 

 

Well buddy , you are making me feel better by saying that all of this is normal though I'm not 100% sure about that I guess I'll have to live with it? Are you certain that these temperature changes I showed in the video are the way the chip operates? I am kind of lost I must say I expected a much different behavior

[faziten]

4 minutes ago, SleVeran said:

Well buddy , you are making me feel better by saying that all of this is normal though I'm not 100% sure about that I guess I'll have to live with it? Are you certain that these temperature changes I showed in the video are the way the chip operates? I am kind of lost I must say I expected a much different behavior

See here:

 

https://www.youtube.com/watch?v=SHebBmuXwsw People all over internet got exactly 90ºC just like you did

 

https://www.youtube.com/watch?v=UuevcySI1BI Just like this one dude.

 

Do a quick reddit search, it's all over the place. You just topped the charts with a bad mix of auto voltage, and prime 95. 

[SleVeran]

42 minutes ago, faziten said:

See here:

 

https://www.youtube.com/watch?v=SHebBmuXwsw People all over internet got exactly 90ºC just like you did

 

https://www.youtube.com/watch?v=UuevcySI1BI Just like this one dude.

 

Do a quick reddit search, it's all over the place. You just topped the charts with a bad mix of auto voltage, and prime 95. 

My computer has been on all night and because my cpu temps are low when it's just doing minor background tasks then my fans aren't spinning fast at all. but if you look at the CPUTIN sensor right now you can clearly see that there is a gap . while my CPU sensor shows 45-40C my CPUTIN (essentially the sensor of the cpu temp provided by the motherboard) gives 70C cause my fans haven't been spinning fast at all.

how can you explain that? 

 

in addition to that , is it normal for my cpu to drop from 4.5Ghz to 800Mhz occasionally? cause it has been happening as well

[faziten]

10 minutes ago, SleVeran said:

My computer has been on all night and because my cpu temps are low when it's just doing minor background tasks then my fans aren't spinning fast at all. but if you look at the CPUTIN sensor right now you can clearly see that there is a gap . while my CPU sensor shows 45-40C my CPUTIN (essentially the sensor of the cpu temp provided by the motherboard) gives 70C cause my fans haven't been spinning fast at all.

how can you explain that?

Just like the other half of internet people that had high temperatures in CPUTIN sensor. Bad reading.

 

You started your test at 58ºC in CPUTIN and ended up in 78ºC. Which gives a delta Tº of 20ºC. That's totally fine considering it's a sensor somewhere in the motherboard close to a CPU that went form 0% usage to 100% usage. 

Most of the time CPUTIN gives a bad reading, just check google. Since 2011 people are having CPUTIN sky high for no apparent reason. Check temperatures inside your BIOS. software is not 100% safe in this regard, if any software is to trust then BIOS it is. 

 

Micro freezes when you have your CPU pegged at 100% is just another normal behavior. 

 

Abnormal would be, reboots, shutdowns, throttling, BSOD, WHEA loggers in event viewer, etc.

 

Your PC is fine just stop toasting it with Prime95. It's not even worth it. 

[SleVeran]

19 minutes ago, faziten said:

Just like the other half of internet people that had high temperatures in CPUTIN sensor. Bad reading.

 

You started your test at 58ºC in CPUTIN and ended up in 78ºC. Which gives a delta Tº of 20ºC. That's totally fine considering it's a sensor somewhere in the motherboard close to a CPU that went form 0% usage to 100% usage. 

Most of the time CPUTIN gives a bad reading, just check google. Since 2011 people are having CPUTIN sky high for no apparent reason. Check temperatures inside your BIOS. software is not 100% safe in this regard, if any software is to trust then BIOS it is. 

 

Micro freezes when you have your CPU pegged at 100% is just another normal behavior. 

 

Abnormal would be, reboots, shutdowns, throttling, BSOD, WHEA loggers in event viewer, etc.

 

Your PC is fine just stop toasting it with Prime95. It's not even worth it. 

Well i guess i should just stop bitching about it then  I moved from i3 2120 with r9 270x just a few days ago built this computer as best as i could in regard of cable management etc. I love it , it runs games beautifully but i was just trying to get involved in the more technical side.. 

Thanks a lot for helping me buddy , i really appreciate that.

[FirstOfHisName]

19 hours ago, SleVeran said:

At first I didn't since the H100i V2 comes with preapplied Thermal compound so I just put it on , but since then i found out the mounting bracket on the H100i V2 has some problem that people complained about so i took care of it by taking it off and putting Spacers on it in order to make the bracket and the water block a tight fit. When i did that fix to the bracket i also applied my own new Thermal compound , of course after cleaning the previous one.