PC turns off while playing games

Very similar to a power supply. Download the OCCT program and test it.

kholodewan
Turns off. First of all, the PSU test - you need to look at the voltages under load, preferably with a multimeter.
- Screenshots roll up under the spoiler.
- CPU temperature. Known topic, 85 limit for old FX processors (protection).
- I would start with an external examination, where you need to evaluate how the PSU itself is - whether the capacitors are swollen, the same external examination of the motherboard, wiring for damage (including from overheating).
- Test each element individually to identify the cause of the shutdown. Or come to some kind of conclusions - "in the aggregate" load.

Previously, CPUTIN 84 degrees. The program can lie, the numbers are sometimes inadequate. core temperature in particular. Update everything or take an alternative.

Also with regards to HWMonitor data, is the computer in an idle state, I understand? If so, then you have a drawdown along the 12 Volt line in the block.

Evil in the flash,
I lowered the voltage on the video card by flashing the BIOS if that

01001000 01101001

Evil in the flash
Wait, where do you see the problem? 11.8 is a valid value.

kholodewan wrote:
I lowered the voltage on the video card by flashing the BIOS if that
He is talking about another. About 12 volts. Are you talking about the voltage on the GPU core? Danuvolting through whore.

kholodewan
Here, on the third screenshot, you can see the drawdown on the 12 volt line. 11.14 is no longer acceptable. A deviation of 5% is possible. You have 8%. BP should be in a landfill.

The most interesting thing is that the problem manifests itself in all games, this is DotA tanks the Witcher 3 Apex, except for doom eternal

01001000 01101001
but in idle everything is fine. Can you restore voltage to the gpu? but the pc still turns off in this case

kholodewan
Update all software, inspect and try to calculate under load.
So far, there is a noticeable dip in the 12-volt line on OCCT. I do not know what condition the PC is in - inspection and evaluation have long been required. So are temperatures.
But probably focus on the BP. 12 volts is HDD, processor, video card, fans.
kholodewan wrote:
but in idle time everything is fine. Can you restore voltage to the gpu? but the PC still turns off in this case
. In idle time, there is no such load. Voltage drops are just fixed under load. How long ago the voltage on the GPU dropped, correlate with the beginning of this problem. Usually, the card will experience driver drops or effective frequency drops.

I'll tell you what, my friend. The computer needs to be serviced. And you didn’t write anything on the case, it’s really bad to know things there, including the need to replace thermal paste. Strongly computer howls?

kholodewan
Listen, it's very hard to work like that. You're posting empty screenshots. Tests take longer.
And finally, take hwinfo64, which has all the information you need on one page.
There is aida64, where you can conduct a stability test and check the test boxes individually or in combination.

Finally, inspect and assess the condition of the components and capacitors first. Particularly BP. And these screenshots, stop sawing here - you drive, you will load with HWiNFO64.

See voltages and power supply. This is the first question for him.

01001000 01101001 I
service often, a week ago I cleaned it from dust, changed the thermal paste on the processor, it didn’t change on the video for a couple of months, mx 5 thermal paste. temperature reached 79 - 83

kholodewan
For this series of cards, this is generally the norm, such temperatures. She has a specification limit of about a hundred, or something. Look. Well done, of course, that he took care and did.
now for the rest of the points. In addition to what was noted above, look at the temperature of the processor, it flies to the defense at 85 (it has a limit).
kholodewan wrote:
mx 5 thermal paste.
I'll tell you something else. Monitor temperatures. Anything can happen, even where it was replaced. She can turn into water. This is so, by the way. There were talks about quality. But this is unlikely. If you're sure, then forget it.
That's right, a small remark.

The max safe operating temperature stated by AMD for the 290X is considered to be 95-98C. Maybe I'm wrong.

01001000 01101001
Just saw the message. Well, you seem to have already painted everything well yourself :) Definitely with such a drawdown, the block is for replacement.

01001000 01101001 Will
10 min test be enough?

kholodewan
kholodewan wrote:
PC just shuts down while playing... fx 4320... Am I missing
something, or is the CPU overclocking a bit? - in the network, the frequencies are indicated as 4.0 - 4.1 GHz, and on the screen 4.4.
Spoiler
If overclocked, monitor its temperature under load - better on top of the game (MSI Afterburner, etc.), where the computer turns off.
Example:
in idle time, everything is fine...
In idle time, there is no such load on the hardware as during games, especially if the frame rate is not limited and the hardware is used to the maximum (or almost to it).
Will 10 minutes of testing be enough?
Quite. But if you want to "warm up" the system - you can do more. During the test, it is better not to leave the computer, but to look at the indicators, especially voltages / temperatures.

SonyK_2 wrote:
I don't understand something, or does the CPU have a slight overclock? - in the network, the frequencies are indicated as 4.0 - 4.1 GHz, and on the screen 4.4.
yes, in overclocking, now I'm testing the processor through occt for about 15 minutes

kholodewan
Oh difficult. Let's go in order.
1. You take the HWiNFO64 program, run it, marking "only sensors" at startup. You will have a complete adequate list of indicators for all kits.
2. Run the test (BETTER let's go for everything and everything at once!) And look in parallel with the numbers that deviate - temperature, voltage. Specifically the 12 volt line.
3. In general, pay attention in this table to what you do not like. Share.
So it seems more or less obvious. I wrote to you to check it out. Everything could have been done. In particular, inspect the PSU and its internal state AND the cables - if there are any melting.
A drawdown on the 12 volt line is already dangerous for the HDD, it does not like it. Let's calculate.

10 minutes, conditionally yes - enough. Take a look at the stability. The game doesn't crash right away.