random shut down on i5-6600 (not PSU problem)

[Maj_Kusanagi]

This problem been bothering me for months now: PC started to random shutdown a few months ago (everything suddenly turns off, event viewer only shows kernel power error). It happened first in June, so I reseated the RAM and it didn't happen again until 1 month later in July. Then it became more frequent from August up to now. But the random shutdowns would happen very randomly (could happen few hours after turning PC on, after 1 day usage or even 3-4 days) , sometimes sitting in idle, browsing, gaming, or leaving the PC on overnight. Besides the random shutdowns, there were a few instances where the screen froze during gaming or light usage (keyboard unresponsive, fans still spinning normally) requiring a hard shutdown, with Windows only showing the kernel power error after turning it back on. Rarely after the shutdowns, the PC does not boot up properly (case & mobo light turns on then off without POST, or powering off right after POST beep, HDD light not active yet), requiring me to push the power button a few times before a successful boot into Windows. After removing the GPU and using onboard graphics only, there were a handful of times where the screen turned into a solid color (yellow, red, or green) with system freezing requiring a hard reset. I'm sure it's not a Windows problem since the shutdown also occurred once when I left it in BIOS for hours.

Steps taken with no fix:
-cleaned dust and reseat different RAM slots
-Intel processor diagnostic quick test passed
-memtest passed
-temps show all normal
-tested SSD and HDD no errors
-reverted BIOS settings
-removed RX570 graphics card using only Intel HD graphics
-swapped PSU with a friend's working one

So last week I finally took the PC to a repair shop, where my PSU was tested with a meter (no abnormalities), tested my ram again without issues, ran OCCT power test and the PC would lockup within minutes. He swapped my CPU with his i5-6400 and ran the same test for 18 hours with no problems. The next day he sound convinced that my i5-6600 is the culprit, explaining that the CPU and integrated GPU are stressed at the same time. However I'm not 100% convinced and think it may be a motherboard problem as well. After taking my PC back home, I repeated the same OCCT power test and noticed my CPU would reach 100c and freeze within minutes (requiring hard shutdown). I ran the other OCCT stability and linpack tests for 1+ hours with no problems (temps reach around 70c), so it's only the OCCT power test that would cause such high temps and freezing. At this point I'm not too sure if it's the high test temps causing such error or really a faulty CPU causing these shutdowns. Everything runs fine and smoothly when my PC is on, hope someone can provide some insight.

System specs:
Intel i5-6600, Gigabyte H170M-D3H, Antec NeoEco II 650w, 8GB Adata XPG DDR4-2133 RAM, Crucial 250GB SSD, 1TB HDD, Windows 10 Home 1909, standard case and peripherals

[AndrewG_Intel]

Hello Maj_Kusanagi

Thank you for posting on the Intel® communities.

In order to check this further, could you please confirm the following information?

1- Just to make sure, we understand that at the beginning the issue happened randomly either gaming, browsing, or in idle state but after taking the computer to a PC repair store, now the issue only happens when running the OCCT* power test. If this correct?

2- After removing your processor from the motherboard to test the other CPU (i5-6400), did you check and replace the Thermal Interface Material (TIM, a.k.a Thermal paste)?

The TCASE (the maximum temperature allowed at the processor Integrated Heat Spreader (IHS) for the Intel® Core™ i5-6600 Processor is 71°C and overheating issues may lead to this type of issues.

https://ark.intel.com/content/www/us/en/ark/products/88188/intel-core-i5-6600-processor-6m-cache-up-to-3-90-ghz.html

3- Have you tested the system with minimal configuration? For instance, testing only the motherboard, CPU, power Supply, integrated graphics and only 1 stick of RAM at the time on different memory slots, only boot drive, wired mouse/keyboard and only 1 display? (No other peripherals, adapters, PCI cards or devices connected to the system).

4- Please run a new Intel® Processor Diagnostic Tool test and provide the report. You can download the proper version (32 or 64 bit) from the this link:https://downloadcenter.intel.com/download/19792/Intel-Processor-Diagnostic-Tool

To save the report, once the test is done, click on "File >> View Results File" and attach the .txt file to your reply.

5- Please run the Intel® System Support Utility (Intel® SSU) and attach the report to this thread to gather more details about your system.

1. Download the Intel® SSU and save the application on your computer: https://downloadcenter.intel.com/download/25293/Intel-System-Support-Utility-for-Windows-
2. Open the application, check the "Everything" checkbox, and click "Scan" to see the system and device information. The Intel® SSU defaults to the "Summary View" on the output screen following the scan. Click the menu where it says "Summary" to change to "Detailed View".
3. To save your scan, click Next and click Save.

Note: To upload and attach a file, use the "Drag and drop here or browse files to attach" dialog below the edit box.

Best regards,

Andrew G.

Intel Customer Support Technician

[Maj_Kusanagi]

Hello Andrew,

Thank you for your help, a quick update before I answer your questions below:

Yesterday I ran the Intel Processor Diagnostic Tool Again, this time choosing the burn in option for a longer test. After it reached the CPU Load section, the monitor lost signal (black screen) and keyboard became unresponsive. The PC remained on I could hear the fan still spinning loudly as if still running the diagnostic. I let it remain in this state for around 10mins, nothing changed so I shutdown holding the power button and rebooted normally into Windows again. Today when I turned on my PC it failed to properly boot multiple times: the system would immediately power off after pressing the power (no post), powered off twice right after post beep, and powered off twice moments after reaching the desktop. 

I reseated the CPU and heatsink fan, and did notice there was fresh thermal paste applied from the last shop visit. After reseating and pressing the power button, it failed to POST and motherboard gave 5 long beeps. I checked Gigabyte reference and it indicated CPU failure although some sources say it's memory. So I reseated the RAM using 1 stick and it successfully booted, allowing me to reply now.

While reseating the CPU, I noticed some contact spots are darker than the rest. I have attached the photo below, is this normal or indication of burning?

Please see the answers to your questions below along with attached Intel SSU log and Processor Diagnostics (default test) results. I will run the Processor Diagnostic burn in (long test) again later to see if the system freezes.

1- Just to make sure, we understand that at the beginning the issue happened randomly either gaming, browsing, or in idle state but after taking the computer to a PC repair store, now the issue only happens when running the OCCT* power test. If this correct?

-No, the random shutdown happened again yesterday. After leaving the PC on idle overnight. I turned on the monitor, started watching youtube and the PC powered off within 1 minute. Also note that OCCT power test did not cause instant shutdown, only system and screen freeze.

2- After removing your processor from the motherboard to test the other CPU (i5-6400), did you check and replace the Thermal Interface Material (TIM, a.k.a Thermal paste)?

-Yes, visually confirmed fresh thermal paste was applied after reseating my CPU today.

3- Have you tested the system with minimal configuration? For instance, testing only the motherboard, CPU, power Supply, integrated graphics and only 1 stick of RAM at the time on different memory slots, only boot drive, wired mouse/keyboard and only 1 display? (No other peripherals, adapters, PCI cards or devices connected to the system).

-Yes, I have tested the system with minimal configuration. It has been running on integrated graphics for over a month now.

[Maj_Kusanagi]

Hi Andrew,

Here's a new update to above reply, the CPU successfully passed the burn-in long test in Intel Processor Diagnostic without crashing, attached results below.

Now I'm really perplexed whether this is actually a faulty processor issue or other hardware.

[n_scott_pearson · Accepted solution]

The graphics engine being unable to read its VRAM (which is part of the standard memory) can also cause system resets. Further, if the memory bus(es) lock up (usually as a result of too much noise on the bus), the system's watchdog timer will reset the system to recover. These lockups are the result of noise on the buses. Noise builds up over time as the components age. A system that works with a particular set of DIMMs that overclock the memory buses when the system is new may not work at all once the system has aged a couple of years.

From the description of your problems and the steps you have taken to try and figure out what is going wrong, my experience is telling me that you have bus lockups occurring. Unfortunately, this could be the fault of the DIMMs, the Motherboard's bus implementation (and noise immunity circuitry) or the processor's memory controllers. Testing with a single DIMM is one way to try and eliminate the DIMMs as the source of the problem (though what you have tried so far hasn't necessarily done that). The rest comes down to trial and error replacement of components.

Hope this helps,

...S

[Maj_Kusanagi]

Hi Scott,

Since the problem persisted when running on single RAM sticks, and memtest showed no errors, is it safe to eliminate the RAM as the source of the problem? The chances of both DIMM sticks being faulty should be quite low, and they have never been overclocked.

At this point is it recommended to replace the processor or motherboard first for testing?

[n_scott_pearson]

Didn't you say you had successfully run with a friend's CPU? If this is the case, it makes sense to try replacing the CPU first.
...S

[AndrewG_Intel]

Hello Maj_Kusanagi

We have not heard back from you so we will proceed to close this thread now. We agreed with n_scott_pearson and at this point, the recommendation is to contact the place of purchase for warranty options. Or if this is an Intel® boxed processor, you may contact Intel® Customer Support directly to check warranty status.

Here is the "Contact us link": https://www.intel.com/content/www/us/en/support/contact-support.html#@2

Please make sure to select your country or region and to contact us within your local business Support hours and to make reference to this thread.

If you need any additional information, please submit a new question as this thread will no longer being monitored.

Best regards,

Andrew G.

Intel Customer Support Technician