My pc restarts while playing games

[Bill_Bright]

Sorry, but I am not wrong. You can see by the link if my sig if I know a little about electronics, hardware, and how system cooling should work.

It is, and always has been the case's responsibility to provide adequate cooling for the components inside. And it the designer's or builder's (or user's) responsibility to design, select and/or configure said case properly.

If you have to run with the side panel open to achieve proper cooling, that's the wrong case, or it is configured wrong! If you believe otherwise, then sorry, but that is simply wrong too.

I don't doubt you have configured your case for the best cooling. But I am saying if you have done that, and it still does not properly cool the components you put in there, then that is simply the wrong case for the job! So then the question becomes, who picked that case?

 

[maxfly]

Take the side panel off and point a desk fan into your case. See if the resulting drop in cpu temps helps with the crashes. If not, you know that heat isnt the issue. Rather the psu is the most likely culprit. Try to borrow your brothers psu to test in your rig. If your able to test with his psu successfully you know for certain your psu needs to be replaced. If not its time to start testing your cpu, gpu and mb.
GL!

 

[matosi]

It restarted after 5 minutes on occt power test gpu was around 85 and cores were around 95 during it.

while stress testing the gpu consider giving the machine some cpu work, eg 7-zip integrated benchmark

and as said by others test, fail and repeat

the truth is out there

 

[Intervention]

Sorry, but I am not wrong. You can see by the link if my sig if I know a little about electronics, hardware, and how system cooling should work.

It is, and always has been the case's responsibility to provide adequate cooling for the components inside. And it the designer's or builder's (or user's) responsibility to design, select and/or configure said case properly.

If you have to run with the side panel open to achieve proper cooling, that's the wrong case, or it is configured wrong! If you believe otherwise, then sorry, but that is simply wrong too.

I don't doubt you have configured your case for the best cooling. But I am saying if you have done that, and it still does not properly cool the components you put in there, then that is simply the wrong case for the job! So then the question becomes, who picked that case

Show me 100 different engineers and I'll show you 100 ways of doing things. Everyone has their own opinion on things. I know that a pc with an RTX 3090, RX 6900 XT or any card with a high power consumption will generate higher amounts of heat. Thermal dynamics, right? I have 5 different systems that I use for my YouTube channel. Each has its own purpose. 2 are open bench systems. Both have an RTX 3090 and a RX 6900 XT. In my other 3 systems I have a RTX 2060 and another RTX 3090 and RX 6900 XT. 3 cases are 1 Corsair 4000D, and 2 Fractal Design Meshify 2. The larger cards are in the FD Meshify 2 cases. Both systems operate cooler with the side panel off. Period.

See the thing is, with all due respect, you keep saying how my pc setups are wrong and questioning what I am pointing out as true. You have no clue what my setup looks like, fan types, graphics card models and cooler designs both for CPU and GPU's, and my ambien conditions in the room where I do my videos. So you are assuming a lot here. I was not even going to respond to your last comment as I see you like to be right a lot based on your previous interactions with other folks. You are right and everyone else therefore must be wrong LOL. I say this with a smile in my face because I deal with difficult people often, as I own a small home renovation business.

Your opinion is valued and I appreciate your thoughts on this. However I will continue to believe what I have seen and tested to work best. And if someone is experiencing difficulties, I will also suggest they try what I have done. It works. But you have a good rest of your day, and remember, opinions are like buttholes. Everyone has one

I would suspect the PSU first. When troubleshooting possible hardware issues, you always start at the wall. Is it plugged in? Is it turned on? Are you delivering good, clean stable power to the electronics? Obviously, it is plugged in and turned on. But the only way to determine if your PSU is delivering good, clean, stable power is to swap in a known good PSU and see if the problems persist, or go away. Since EVERYTHING inside the case depends on good power, you must start here.


Doesn't matter. It is a myth (and often a risky one too) that TIM (thermal interface materials) needs to be replaced because it is X number of years old. Folks really should just leave TIM alone. AS LONG AS the cured bond between the processor and the heatsink is not broken, there is no reason to replace it. Even if it is 10 years old and dried out, the solids left behind are still occupying the microscopic pits and valleys in the mating surfaces, preventing insulating air from getting in. In other words, the TIM is still doing its job. If you truly "need" the few degrees a fresh application of new TIM "might" provide to prevent crossing over thermal protection thresholds, you have other, more urgent heat issues to deal with first, like case cooling.

I note there is not a single TIM maker, motherboard maker, CPU or GPU maker, or cooler maker that recommends or even suggests replacing the TIM after X number of years.

The problem is, it is too easy to have a mishap that damages the processor, causes excess TIM to get on where it should not be, causes the heatsink to be improperly remounted, etc. All for something not needed. Best to leave it alone so that cured bond does not get broken.


That's fine, Contrary to what many want us to believe, OEM/stock coolers of today are more than capable of providing adequate cooling for the processors they came with AS LONG AS the case is providing a sufficient supply of cool air flowing through the case. So make sure the case is clean of heat trapping dust and your fans are spinning properly.

I note HWiNFO does show some extremely hot temps, but they are so high (127°C), most likely they are false readings. You might see what another monitor, like Speccy, says.

Not true. Noctua recommends "upto" 5 years usage on a CPU for their NT-H2 thermal pastes. If they recommend up 5 years, what does that mean to anyone? What can anyone draw out of that statement? Logically, I think, it means you should replace it after 5 years. But again, I know you will say I am wrong LOL

 

[Arvellon]

Take the side panel off and point a desk fan into your case. See if the resulting drop in cpu temps helps with the crashes. If not, you know that heat isnt the issue. Rather the psu is the most likely culprit. Try to borrow your brothers psu to test in your rig. If your able to test with his psu successfully you know for certain your psu needs to be replaced. If not its time to start testing your cpu, gpu and mb.
GL!

Sadly we don't have any backup PSU we can test since my brother is using a laptop and it started to get worse after the cleaning pc started rebooting even with league so there is that as well.

Hello. Have you tried running everything with your computer side panel off? This to allow your computer hardware to breath a bit better and allow for hot air to dissipate into the air quicker. Your cpu and gpu should run a bit cooler. If the symptoms stop then your problem is most likely due to overheating. Try reapplying your thermal paste for cpu and gpu.

Sadly taking side panel off didn't help reapplying thermal paste did make temperatures drop for cpu and gpu but it still reboots.

 

[Intervention]

Sadly we don't have any backup PSU we can test since my brother is using a laptop and it started to get worse after the cleaning pc started rebooting even with league so there is that as well.


Sadly taking side panel off didn't help reapplying thermal paste did make temperatures drop for cpu and gpu but it still reboots.

Okei. Have you tried memtestpro to test memory reliability? Also, a Prime95 small ffts? 1 hour should do. This just to rule out memory or cpu weaknesses.

 

[Arvellon]

Okei. Have you tried memtestpro to test memory reliability? Also, a Prime95 small ffts? 1 hour should do. This just to rule out memory or cpu weaknesses.

I couldn't download prime95 anywhere so I couldn't test that.

 

[eidairaman1]

SFC

 

[basco]

i think eidaira means this:

Use the System File Checker tool to repair missing or corrupted system files - Microsoft Support

Describes how to use the System File Checker tool to troubleshoot missing or corrupted system files in Windows 8.1, Windows 8, Windows 7 or Windows Vista.

support.microsoft.com

 

[eidairaman1]

i think eidaira means this:

Use the System File Checker tool to repair missing or corrupted system files - Microsoft Support

Describes how to use the System File Checker tool to troubleshoot missing or corrupted system files in Windows 8.1, Windows 8, Windows 7 or Windows Vista.

support.microsoft.com

Yes that is correct

 

[Intervention]

I couldn't download prime95 anywhere so I couldn't test that.

You can download it from Techpowerup: https://www.techpowerup.com/download/prime95/

 

[Bill_Bright]

Show me 100 different engineers and I'll show you 100 ways of doing things. Everyone has their own opinion on things.

It is NOT about opinion. It is about fact. A computer case has two primary responsibilities. (1) To house and protect from physical abuse (kicks and bumps) the components inside. And (2), to provide sufficient cooling for those components inside.

If you have to remove the side panel to keep the components inside properly cooled, that case is NOT doing its designated job. That is just a fact - not opinion. It could mean that it is the not correct case for the job, or the case's cooling is configured wrong. Either way, the case is not properly cooling the components. And selecting the right case for the job, and configuring case cooling properly is the user's (or builder's) responsibility. Again, not opinion, just a fact.

If you buy a brand new Ford F150 pickup truck, then hook up a 15,000 pound trailer to it, and the truck overheats, then YOU bought the wrong truck for the job. You should have bought a bigger truck or got a lighter trailer.

Why do cases have air filters if you have to remove the side panel?

As far as Noctua recommending "up to" 5 years usage - LOL, nice try!

You are right -I will say that is not the same thing as Noctua recommending changing it every 5 years. And that's because they don't recommend it. Why? Because you don't have to! Had you dug just a little bit deeper, you would have seen here where it says, my bold underline added,

NT-H2 unique formula is highly stable over time...it can be used on the CPU for 5 years or more.

So no, that does not mean you "should" replace it every 5 years, as you "concluded". It means what it says - it will easily "remain stable" at least that long. So if I were to "conclude" anything by that, it would be that the chemical composition of the thermal compound will not break down for "5 years or more". And, as I noted repeatedly above, as long as the cured bond is not broken, thus preventing air (or in particular, oxygen) from getting inside, there's really no reason for the chemical composition to break down - thus allowing the TIM to remain effective for years to come.

Ask around with the old-timers here and elsewhere. No doubts many have seen computers (and other electronics) that have not been touched for 10 years (or even longer), still chugging along just fine. Not just with the same original application of TIM, but with the same OEM HSF assembly and, astonishingly (to me anyway) even the same, cheap PSU!

Once again, I am not denying (and never have) that a fresh application of TIM may provide a few degrees of improved cooling. I am simply saying if you are that close to critical thresholds that a few degrees matters, you have bigger cooling issues to deal with first.

 

[Arvellon]

Yes that is correct

It didn't find anything.

Okei. Have you tried memtestpro to test memory reliability? Also, a Prime95 small ffts? 1 hour should do. This just to rule out memory or cpu weaknesses.

I attached the results text it crashed after 15 minutes or so.

 

[Intervention]

It is NOT about opinion. It is about fact. A computer case has two primary responsibilities. (1) To house and protect from physical abuse (kicks and bumps) the components inside. And (2), to provide sufficient cooling for those components inside.

If you have to remove the side panel to keep the components inside properly cooled, that case is NOT doing its designated job. That is just a fact - not opinion. It could mean that it is the not correct case for the job, or the case's cooling is configured wrong. Either way, the case is not properly cooling the components. And selecting the right case for the job, and configuring case cooling properly is the user's (or builder's) responsibility. Again, not opinion, just a fact.

If you buy a brand new Ford F150 pickup truck, then hook up a 15,000 pound trailer to it, and the truck overheats, then YOU bought the wrong truck for the job. You should have bought a bigger truck or got a lighter trailer.

Why do cases have air filters if you have to remove the side panel?

As far as Noctua recommending "up to" 5 years usage - LOL, nice try!

You are right -I will say that is not the same thing as Noctua recommending changing it every 5 years. And that's because they don't recommend it. Why? Because you don't have to! Had you dug just a little bit deeper, you would have seen here where it says, my bold underline added,



So no, that does not mean you "should" replace it every 5 years, as you "concluded". It means what it says - it will easily "remain stable" at least that long. So if I were to "conclude" anything by that, it would be that the chemical composition of the thermal compound will not break down for "5 years or more". And, as I noted repeatedly above, as long as the cured bond is not broken, thus preventing air (or in particular, oxygen) from getting inside, there's really no reason for the chemical composition to break down - thus allowing the TIM to remain effective for years to come.

Ask around with the old-timers here and elsewhere. No doubts many have seen computers (and other electronics) that have not been touched for 10 years (or even longer), still chugging along just fine. Not just with the same original application of TIM, but with the same OEM HSF assembly and, astonishingly (to me anyway) even the same, cheap PSU!

Once again, I am not denying (and never have) that a fresh application of TIM may provide a few degrees of improved cooling. I am simply saying if you are that close to critical thresholds that a few degrees matters, you have bigger cooling issues to deal with first.

It didn't find anything.

I attached the results text it crashed after 15 minutes or so.

So it crashed after 15 minutes of you running prime95 small fft's, correct? I would try this: Reset the BIOS. Do not enable XMP. Run everything stock. See if the system is stable. Start there. If the issue goes away, then perhaps some voltage setting got out of wack. Enable XMP and see if system continues to work without issue.

If the issue continues. Reset your RAM physically, your graphics card, and check that all PSU connections to the motherboard are set correctly. I'd unplug and plug them back in just to reset all.

 

[eidairaman1]

Set chipset voltages a little higher than factory default in bios. May need to go to a shop to swap parts for testing... All parts can cause a restart, its a matter of isolating it.

 

[Arvellon]

So it crashed after 15 minutes of you running prime95 small fft's, correct? I would try this: Reset the BIOS. Do not enable XMP. Run everything stock. See if the system is stable. Start there. If the issue goes away, then perhaps some voltage setting got out of wack. Enable XMP and see if system continues to work without issue.

If the issue continues. Reset your RAM physically, your graphics card, and check that all PSU connections to the motherboard are set correctly. I'd unplug and plug them back in just to reset all.

We already did all of that including unplugging and plugging them back and it didn't help.

 

[Bill_Bright]

May need to go to a shop to swap parts for testing...

^^^This^^^. If a second, known good PSU didn't resolve this, and you are sure all connections are tight and secure, I might try running with just one stick of RAM at a time.

 

[DrCR]

@Arvellon, how did things end up for you?

 

[Arvellon]

@Arvellon, how did things end up for you?

Sadly nothing helped it looks like I'm going to have to change PSU.