Turning off or freezing the computer tightly in games - how to determine if the reason is in windows or hardware?

Question

Iron

Configuration
Processor: i5-2500k (no overclocking)
Motherboard: Pegatron IPMSB/H61/ODM
RAM: 6GB
Video card: Gygabite 1060 3GB.
Hard drive - I don't remember the exact model, 1TB, 5400 rpm.
System: windows 7
The computer sometimes restarts or freezes tightly in games.
As a rule, this happens in some + - one place in the game, after several cases it goes away.
The last time it happened, it started to turn off at the same level, many times the
temperatures are completely normal (do not exceed 60-65 degrees), checked with MSI Afterburner.
I checked the memory using memtest86 - I did not find any problems, the hard drive using victoria hdd - the same picture, just one green sector, there are no red and yellow ones at all.
I ran the processor benchmark and furmark at the same time - the processor and video card were almost completely loaded, there were no crashes, I concluded that it was not a matter of the PSU, processor or video card (the games in which it crashes can be quite old and not able to fully load the hardware (fallout 3 NV for example))
In the system log, each shutdown comes with a kernel-power 41 error.
The system is three years old, survived the motherboard change.
If the matter is in the system, then what is happening in it, why is it turning off?
Thanks in advance

Answer 1

1. Memory. Remove, wipe, clean the contacts, rearrange the dima in some places. Electronics is the science of contacts. If there is anything, replace it. If there is money - to expand (the bug will not disappear, but it may become less common)
2. Video. For about half a year I could not understand where the glitch came from - in games or with the equipment test built into Windows, the monitor suddenly goes out, the video screams like a liner taking off. I changed the vidyuhi - the glitch disappeared (it's curious that this vidyukha is now plowing in another computer).
A sudden instant freeze is usually a piece of iron, not software.

Answer 2

How to check that the matter is in the system - elementary - reinstall the fucking system. For three years in the seven, a lot of shit can accumulate, especially if the iron was changed.

Answer 3

Eliminating a critical Windows Kernel Power error, event code 41
On the other hand, even the "Microsoft" source does not explain such a failure. Kernel-Power (event ID 41) in the form of a failure can occur for a variety of reasons. These are already enthusiasts of their business literally got to the bottom of possible symptoms. The saddest thing is that they differ so much that finding out the real cause of the failure is far from the easiest thing. In addition, you can find an indication of such a failure only in the reports (visual confirmation is not issued).

No cause of disease - no cure? After the computer or laptop stops responding to user manipulations, only a full Reset (reboot) helps, but even then, only by long pressing and holding the power button. By the way, it is far from a fact that after that the system will work normally. Some time - maybe, but then the re-occurrence of the Kernel-Power error (event code 41) is absolutely not excluded. And, judging by the fact that even Microsoft experts cannot name the true cause of the failure, many users are unable to determine its true nature. So what to do in this case?
Separately, it should be said that reinstalling Windows does not help either. So from here we can draw the only possible conclusion: the conflict is at the level of hardware, system and installed software. However, in the case of programs or games, the explanation of the failure by the fact that the hardware simply does not meet the recommended system requirements looks quite logical. However, despite the lack of official transcripts, as is customary for most Windows errors, there are still several treatment options. Well, firstly, you should check whether all the cables fit snugly, and secondly, the power supply (if any) can simply “not pull” the load, causing power surges. But now is not about that.

Kernel-Power critical error (event code 41): setting up the BIOS The most common failure occurs when the processor overheats. This can be observed in two cases: either it is under load that it is not designed for in the software plan (resource-intensive games), or it was simply overclocked.
To begin with, if a Kernel-Power problem occurs, event code 41 (Windows 7, for example), you should look at the characteristics of the central processor and change all overclocked parameters in the BIOS that correspond to overclocking (we are not talking about physical overclocking now). For an ordinary user, in order not to delve into the settings, and without special knowledge, the easiest way is to reset the settings to the factory settings.
As a rule, in any BIOS there is an option Load BIOS Setup Defaults or something like that. Don't like the software way?
In stationary computers on the motherboard, you can remove the battery for a few minutes or move the Clear CMOS jumper from the “1-2” position to the “2-3” position for 15 seconds, then return everything to its place. As already clear, the settings will be reset. However, with a weak processor (especially with its physical overclocking), this may not always help. Yes, and please note: this method is only suitable for desktop computers. For laptops, it is absolutely unacceptable.

In the simplest version, you can use the free utility Everest or something similar. True, it only tests hardware components, and certainly does not restore the original or optimal parameter values. But, at least, at least it will be known to what extent such a failure can be related precisely to overheating of the processor.
In general, it is best to use the so-called Prime95 test, where you first need to select the Just Stress Testing mode, and set the options in the Torture Test ... section to the In-place large FFTs mode.
In this situation, for starters, you can use the standard "RAM" checker, which is present in any Windows system. You can access it, for example, by entering the appropriate query in the Start menu search field or by entering the mdsched command in the Run menu, and then select the desired action (reboot with verification or test at the next system boot). Please note: in both cases, the utility should be launched as an administrator, and if necessary, enter a password, if one has been set. On the other hand, the test may not reveal anything critical, and the user will again get the Kernel-Power (41) error. What to do in such a situation? It's simple - you just need to take out the memory strips one by one and then reboot the system and see how stable it works. Maybe, the problem is in the recently installed slats or simply in their inconsistency with each other. Again, this can be done on stationary computer terminals.
In this case, it is recommended to use utilities like HDD Life or HDD Health and look at the test results. In some cases, it may be necessary to update the firmware of the hard drive, although this is very rare. If testing shows that there are problems with the "screw", you will either have to carry it in for repair, or, alas, install a new one.
In extreme cases, you can try the HDD Regenerator software package, which, according to the developers, can not only repair bad sectors, but even remagnetize hard drives of any type. However, there is no full confidence in restoring their performance, so you will have to use this utility at your own peril and risk.
As it is already clear, they must either be removed first and then reinstalled, or simply updated. A standard system tool or even a free version of the Driver Booster utility, which is able to perform all actions automatically in a few minutes, is perfect for updating. If you do not have an Internet connection, you will have to download the latest drivers from another computer and then install them manually.

Be sure to turn on Windows Updates Finally, many experts who have investigated this failure agree that it is mandatory to turn on automatic system updates. This can solve many problems related to the system side, since the updates practically do not affect the hardware segment (with rare exceptions in the form of updating specific drivers).

Answer 4

Open the system unit, blow out the dust, change the thermal paste, check the capacitors for swelling.