[SOLVED] i7-4790k is overheating.

Jul 20, 2019
7
0
10
Hi,

lam using i7 4790K processor for more than 4 years. It was
working fine till last week without any issue.

Before 2 days my system suddenly went down. I got doubt and
installed H/W monitoring tools to check abnormalities. During
this I found CPU temperature is spiking to 100"c even at 40%(At ideal it's between 40° to 45°)
of load i tried installing multiple temp monitoring tools all are
showing the same result.

Here I want to highlight one more thing that cpu temper erature at bios is higer than the readings at os level tools.

I tried cleaning my system and applied new thermal paste, updated UEFI firmware to the latest, disabled speedstep and Intel turbo boost technology but issue still persist.

Below are my system configuration:-

Processor:- I7 4790K
Mobo:- Asrock z87 extreme4
PSU:-Corsair RM1000
Cooler:- Corsair H100i
SSD:- 250 GB Samsung 850 pro
HD:-4TB(2TB of Seagate and 2TB of WD)
RAM:-16 Gb Corsair vengence 1600
Graphics card:- Asus strix GTX1070
 
Solution
People do not always detect issues with liquid coolers...

Contaminants, hard water deposits, and micro-leaks of fluid levels over the years all combine to make the cooler eventually 'substandard'...

It is not always a rock-solid/easy to see 'my pump stopped pumping' sort of affair...

It might have degraded slowly, but, most don't notice it until proverbially 'smacked in the face'with obvious symptoms of actual CPU clock speed throttling...

A CPU suddenly beginning to overheat is quite often an obvious symptom and consequence of using an AIO...; unfortunately, the AIOs are becoming needed these days due to the some of the '95-105 watt' (cough-150-200 watt!) mainstream behemoths.
Jul 20, 2019
7
0
10
Hi, I'd blame the cooler pump. Try testing different cooler.
I installed Corsair lynk4 to check the cooler performance and I don't find any issues with the cooler as per tools readin. Even I tried replacing my cooler with Intel stock cooler in an air conditioned room with temperature of 22°c(aprox), now the ideal temperature dropped to 29°c but when I increased the load to 40% again it went to 100°c.
 

CompuTronix

Intel Master
Moderator
Ravinovia,

On behalf of Tom's Moderator Team, welcome aboard!

In reference to alexoiu's comment, there's nothing wrong with your CPU.

All 4 individual Core temperature sensors (Digital Thermal Sensors) just don't suddenly and simultaneously all fail. It's extremely rare for even a single thermal sensor to fail.
  • Your H100i has failed.
It happens all too frequently. We see it here in our Forums numerous times every day. Core temperatures that suddenly go high indicate a flow problem. Pump RPM in BIOS or a software utility does not prove proper flow. The impeller can still rotate with a blockage, which is very common.

When you installed the stock cooler you inadvertently introduced a second problem. This created a misleading result which is masking the original problem, thus causing confusion.
  • Your Intel stock cooler has a loose push-pin.
The push-pins are deceivingly tricky to get them all completely inserted through the motherboard and securely latched without one of them popping loose. Even the most experienced of us have all made that mistake. This causes poor contact pressure where idle may seem relatively normal, but even a moderate load immediately goes to Throttle temperature at 100°C.

Reinstall the stock cooler, but use a strong light with attention to detail. Closely examine each pin and compare it to its neighbor. If your case has a motherboard tray with a cut-out that shows the back of the socket, you can easily check each pin. Once you get the stock cooler properly installed and secured, it'll be fine for gaming etc, as long as you don't run extremely heavy workloads.

Corsair has a 5 year warranty, so you can simply RMA the H100i at your leisure ... OR ... you can instead choose to replace it with a decent air cooler. Just FYI, all AIOs will fail ... it's not a question of if, it's a question of when, so having a backup close at hand is always a good idea. Air coolers are fairly bullet proof ... IF they don't have push-pins. There's no substitute for proper attachment hardware with a backplate.

Further, I would recommend that you permanently install a simple monitoring utility such as Core Temp. It's accurate, has very light resource usage, starts with Windows and runs in the Tray so you're always aware of your Core temperatures.

Note: Your i7-4790K is an 88 Watt TDP processor for which Intel specifies their 95 Watt TDP cooler. Many users mistakenly install Intel's 65 Watt TDP version from their previous i5, i3 or Pentium computer. At first glance both aluminum coolers look identical, but when viewed from the bottom, the 95 Watt version has a large copper insert in the center, while the 65 Watt version is all aluminum.

Once again, welcome aboard!

CT :sol:
 
CompuTronix has a point. But, in case his possible causes show up to be false, there is one last possibility: the thermal paste inside the CPU has cooked itself.
After you've gone through all of CompuTronix's suggestions, you should add de-lidding the CPU to the list.
 
Jul 20, 2019
7
0
10
mitch074,

That's a fair thought, however, Intel's internal TIM fails gradually over a lengthy period of time such as in years. In contrast, the OP's problem appeared seemingly within a week, so the symptom doesn't fit the scenario.

Thanks Compu Tronix and mitch074 for your suggestions. Let me test this and come back to you.
 
Jul 20, 2019
7
0
10
I tried as you
Ravinovia,

On behalf of Tom's Moderator Team, welcome aboard!

In reference to alexoiu's comment, there's nothing wrong with your CPU.

All 4 individual Core temperature sensors (Digital Thermal Sensors) just don't suddenly and simultaneously all fail. It's extremely rare for even a single thermal sensor to fail.
  • Your H100i has failed.
It happens all too frequently. We see it here in our Forums numerous times every day. Core temperatures that suddenly go high indicate a flow problem. Pump RPM in BIOS or a software utility does not prove proper flow. The impeller can still rotate with a blockage, which is very common.

When you installed the stock cooler you inadvertently introduced a second problem. This created a misleading result which is masking the original problem, thus causing confusion.
  • Your Intel stock cooler has a loose push-pin.
The push-pins are deceivingly tricky to get them all completely inserted through the motherboard and securely latched without one of them popping loose. Even the most experienced of us have all made that mistake. This causes poor contact pressure where idle may seem relatively normal, but even a moderate load immediately goes to Throttle temperature at 100°C.

Reinstall the stock cooler, but use a strong light with attention to detail. Closely examine each pin and compare it to its neighbor. If your case has a motherboard tray with a cut-out that shows the back of the socket, you can easily check each pin. Once you get the stock cooler properly installed and secured, it'll be fine for gaming etc, as long as you don't run extremely heavy workloads.

Corsair has a 5 year warranty, so you can simply RMA the H100i at your leisure ... OR ... you can instead choose to replace it with a decent air cooler. Just FYI, all AIOs will fail ... it's not a question of if, it's a question of when, so having a backup close at hand is always a good idea. Air coolers are fairly bullet proof ... IF they don't have push-pins. There's no substitute for proper attachment hardware with a backplate.

Further, I would recommend that you permanently install a simple monitoring utility such as Core Temp. It's accurate, has very light resource usage, starts with Windows and runs in the Tray so you're always aware of your Core temperatures.

Note: Your i7-4790K is an 88 Watt TDP processor for which Intel specifies their 95 Watt TDP cooler. Many users mistakenly install Intel's 65 Watt TDP version from their previous i5, i3 or Pentium computer. At first glance both aluminum coolers look identical, but when viewed from the bottom, the 95 Watt version has a large copper insert in the center, while the 65 Watt version is all aluminum.

Once again, welcome aboard!

CT :sol:
Hi Compu Tronix,

I tried removing the stock cooler and reinserted it, this time I made sure that all the push pins are properly mounted as you mentioned but still issue remains same. I checked the stock cooler it's having copper at it's bottom as you described so I guess I am using the correct stock cooler.
 

CompuTronix

Intel Master
Moderator
Ravinovia,

As per my original statement, SgtScream is correct:
  • Your H100i has failed.
Nonetheless, I will say it again:
  • You still have a loose push-pin ... you just haven't found it yet.
While I agree that the stock cooler is inadequate for the i7-4790K, it will work well enough on auto BIOS settings ... IF it's installed correctly, and IF you're not overclocked, and IF you're not running heavy workloads. If it's not installed correctly, then just as SgtScream also stated, it will go to 100°C, even when under light workloads.

As I previously mentioned, since your H100i is still covered by Corsair's 5 year warranty, you should take advantage of that and RMA the unit for a replacement. You can then decide if you instead want a decent air cooler that has a back plate ... but no push-pins.

In the interim, if you don't want your rig to be down, then try again to remount the stock cooler.

Installing Intel's stock cooler:


When you're finished, clear CMOS before you power up, which will reset BIOS to default Vcore settings. Your motherboard manual shows how to do it.

CT :sol:
 
Jul 20, 2019
7
0
10
I've built many 4th gen i7 systems and I can tell that the stock cooler will not be able to keep temps under 100c. That being said, your pump is shot on the AIO. Get a higher quality air cooler that will fit your case. What computer case do you have so I can suggest a cooler?
Hi SgtScream,

I am using NZXT Phantom 410 mid tower case.
 
People do not always detect issues with liquid coolers...

Contaminants, hard water deposits, and micro-leaks of fluid levels over the years all combine to make the cooler eventually 'substandard'...

It is not always a rock-solid/easy to see 'my pump stopped pumping' sort of affair...

It might have degraded slowly, but, most don't notice it until proverbially 'smacked in the face'with obvious symptoms of actual CPU clock speed throttling...

A CPU suddenly beginning to overheat is quite often an obvious symptom and consequence of using an AIO...; unfortunately, the AIOs are becoming needed these days due to the some of the '95-105 watt' (cough-150-200 watt!) mainstream behemoths.
 
Solution