[SOLVED] Screen freezes, pc keeps running(unresponsive) and requires hard shutdown.\\\

[Backhander]

I have bought this PC 3 years ago off of a FNAC store, it was already fully built and game ready. Then like 6 months after the purchase is when i got the first ever freeze/shut down. It kept happening every 2-3 days or sometimes once a week, i just long pressed the power button and turned it on and ignored the problem, i was kinda lazy, altho it never happened during gaming, always when pc was idling or if i was watchin youtube/netflix etc. It always like this, untill i installed the new WoW Shadowlands( 2 months ago) and whenever i went into a scenario where there was multiple online players 15-20 minutes in, i would get the same kind of shut down/freeze. It really started to annoy me so i asked my friend that knows a little more about PCs whats the problem, he told me to check the temperatures and clean the dust inside the build. I've found out that my graphics card is running at 97 celsius at any game i play(league of legends, cod warzone, pubg,valorant and even WoW Shadowlands), the thing is, when i run all these games(except Shadowlands) it runs on stable 144frames with no real problems or lags, but when i run Shadowlands, its soooo poor and even my graphics settings are the lowest they can be, the game runs on 20 frames basically for good 10 minutes and then it simply freezes my pc and i have to shut it down. What i did is i opened the side panel of my pc for the first time since i bought it, i cleaned it properly and carefully, same stuff was happening, and i was getting same temperatures, i checked if all fans were working properly and i even opened the window that the pc is sitting next to and there was no change in the temperature or the pc crashes. So then i went and bought highest quality thermal paste and applied them properly (by watching youtube tutorials). Same stuff, temp literally didnt change and i really dont know what else to do or think, i have come to think that the GPU might not be the problem at all anymore but its something else, i tried everything that internet could offer except posting on a site like this, so this is my last try to save this PC.
Also, note that i've never overclocked this PC, in fact i tried underclocking it and it still didnt work.

Here's my PC specs:
CPU - Intel Core i5 8400 @ 2.80 GHz
Motherboard - Acer Predator PO3-600
RAM Memory - 2x8GB DDR4
GPU - NVIDIA GeForce GTX 1060 6GB GDDR5
PSU - LiteOn 7501 500W

 

[RARRAF]

No PSU info make and model and rough age normally the problem.

 

[Backhander]

No PSU info make and model and rough age normally the problem.

PSU is LiteOn 7501 500W

 

[AdamG]

I've found out that my graphics card is running at 97 celsius at any game

https://www.geforce.com/hardware/desktop-gpus/geforce-gtx-1060/specifications
Thermal and Power Specs
Maximum GPU Temperature (in C)
94

Could be throttled from the temps however it should be noted its reference design thermal limit. I don't think its neccesarily a cause for the pc to "freeze", typically I wouldn't expect it to cause system instability considering the throttle features of the card though.

 

[Backhander]

https://www.geforce.com/hardware/desktop-gpus/geforce-gtx-1060/specifications
Thermal and Power Specs
Maximum GPU Temperature (in C)
94

Could be throttled from the temps however it should be noted its reference design thermal limit. I don't think its neccesarily a cause for the pc to "freeze", typically I wouldn't expect it to cause system instability considering the throttle features of the card though.

Note that it never goes over 97 but also never goes under 97 when gaming, when idling, goes down to 50-60C

 

[AdamG]

Note that it never goes over 97 but also never goes under 97 when gaming, when idling, goes down to 50-60C

97c is not a typical temp for a 1060 under load, I run a 1060 and it doesn't even hit 80c under full load. you stated you checked the all of the fans were running, so I'm assuming you checked the GPU fans also? Are your case fans pulling in air in the front/top and exhausting in the back, and not potentially backwards on the intake/exhaust? since you stated your running at 97c im assuming you pulled that temp from a program like hwmonitor? maybe you should consider msi afterburner to attempt to set a more aggressive fan curve at lower temps, but with 97c id assume your card is running 100% fan before it even hits 90, and still increases in temps after, not typical at all.

I've checked around and seen the pc you are referencing is a pre-built available on numerous sites, so I would assume the fans were installed correctly. The PSU isn't exactly one most would consider a quality psu, reference the PSU tier list. I would surely address the temp issues before worrying about your PSU though, rather then compounding the issues and ignoring it. This build you purchased comes with "PredatorSense Utility App"? Could you perhaps turn case fan rpm up through that utility, and also as previously asked, you verified the gpu fans are running, maybe check their fan curve in msi afterburner or something.

50c idle isn't super abnormal, I'm idling around 40c without load, so its not that far off, but I would check the utility app and see if there is any potential to increase case fan rpm if possible. I wouldn't expect temps to drop by opening case, as you decrease the fans ability to move air through the case by opening it.

 

[Backhander]

97c is not a typical temp for a 1060 under load, I run a 1060 and it doesn't even hit 80c under full load. you stated you checked the all of the fans were running, so I'm assuming you checked the GPU fans also? Are your case fans pulling in air in the front/top and exhausting in the back, and not potentially backwards on the intake/exhaust? since you stated your running at 97c im assuming you pulled that temp from a program like hwmonitor? maybe you should consider msi afterburner to attempt to set a more aggressive fan curve at lower temps, but with 97c id assume your card is running 100% fan before it even hits 90, and still increases in temps after, not typical at all.

I've checked around and seen the pc you are referencing is a pre-built available on numerous sites, so I would assume the fans were installed correctly. The PSU isn't exactly one most would consider a quality psu, reference the PSU tier list. I would surely address the temp issues before worrying about your PSU though, rather then compounding the issues and ignoring it. This build you purchased comes with "PredatorSense Utility App"? Could you perhaps turn case fan rpm up through that utility, and also as previously asked, you verified the gpu fans are running, maybe check their fan curve in msi afterburner or something.

50c idle isn't super abnormal, I'm idling around 40c without load, so its not that far off, but I would check the utility app and see if there is any potential to increase case fan rpm if possible. I wouldn't expect temps to drop by opening case, as you decrease the fans ability to move air through the case by opening it.

First of all, thanks for taking your time to help me with this problem! Yes, indeed i checked the gpu fan(there is only one) and its working properly same as the case fans(sucks air in from the front and exhausting in the back.) I checked temps on various programs (MSI Afterburner being one of them) it always shows 97c when gaming on all of them. The graphics card fan curve is set to "Auto" in MSI Afterburner but however its almost always on 100%.

Yes i tried running the PredatorSense app, it shows your temps and lets you control the case fan RPM. When i first opened the app the RPM was sitting on auto but never moved above 1333RPM so i went and put it on highest RPM possible (4000) and played games for a day or two. Result is the same temps and same crashes on WoW Shadowlands or when idling, so i figured i would turn it down cuz it was just unnecessarily loud.
Note that i started my pc 15 minutes ago and went straight forward to this site to see some answers and as i was typing this reply, i got the same kind of freeze/crash while my GPU was sitting on 49C... Also i should state that my CPU is sitting between 30-60C when gaming or idling, never above.

 

[AdamG]

First of all, thanks for taking your time to help me with this problem! Yes, indeed i checked the gpu fan(there is only one) and its working properly same as the case fans(sucks air in from the front and exhausting in the back.) I checked temps on various programs (MSI Afterburner being one of them) it always shows 97c when gaming on all of them. The graphics card fan curve is set to "Auto" in MSI Afterburner but however its almost always on 100%.

Yes i tried running the PredatorSense app, it shows your temps and lets you control the case fan RPM. When i first opened the app the RPM was sitting on auto but never moved above 1333RPM so i went and put it on highest RPM possible (4000) and played games for a day or two. Result is the same temps and same crashes on WoW Shadowlands or when idling, so i figured i would turn it down cuz it was just unnecessarily loud.
Note that i started my pc 15 minutes ago and went straight forward to this site to see some answers and as i was typing this reply, i got the same kind of freeze/crash while my GPU was sitting on 49C... Also i should state that my CPU is sitting between 30-60C when gaming or idling, never above.

Not a problem, I am merely suggesting things I would try personally. I have multiple spare power supplies, video cards (1060s/650s) and a backup board/cpu so I can't really suggest everything I would try as I am sure you are limited to the rig you have only. The symptoms appearing without GPU temps being high is troublesome though. I pulled this from nvidia website:
https://nvidia.custhelp.com/app/ans...maximum-operating-temperature-and-overheating

"If a GPU hits the maximum temperature, the driver will throttle down performance to attempt to bring temperature back underneath the maximum specification. If the GPU temperature continues to increase despite the performance throttling, the GPU will shutdown the system to prevent damage to the graphics card. "

So it will shutdown the system to prevent damage to the card if the temp increases past maximum; that doesn't really answer why you would get a crash without the temp being abnormally high. As mentioned before the PSU could be a contributing factor to your system instability, a rail could be dead but the psu still will turn on. Personally I would go through testing ram just to be sure, substitute out the psu for a better quality psu (testing kits for psu are about 15$ though) and test another gpu as that one is reaching the point of interrupting system stability. I would be starting with the ram first as it would be easiest for me to test (not relating to gpu temps), then for me replacing the gpu would obviously be a requirement as it shouldn't come close to 97c even under full load, especially with proper airflow, but for you, you could always increase gpu fan curve to have the fan at 100% once card hits 65c if you wanted to attempt to see if that makes any significant difference, doubtful it will though, you could lock the clock at a lower frequency too, like 1600/1700ish if you desired to try that.
Locking GPU clock (Youtube)
Setting more aggressive fan curve (Youtube)

I can't help you much with mini dumps (assuming you have actual blue screen or kernel dumps to share after enabling the .dmp files). That is the only reason I would personally go through testing ram with memtest to make sure that was not contributing factor to the freeze/crash without card being high temps, and its free/easy to do and removes that as a factor from the issues leaving the psu/gpu as likely source(s) of the problem. The PSU/GPU issue will most likely cost money to avoid the symptoms, if you come to the reasonable conclusion it is the cause of your crash/freeze. I'm not personally confident the GPU is the cause of all your crashes/freezes, but certainly hitting 97 you would reasonably believe its going to shut down to avoid damage to the card.

PSUs are not terribly expensive either, you don't need like 80 plus platinum rated one for a 1060 n i5-8400 anyway its a lower-mid tier system, so I don't really see any issue with looking into one of the 80 plus bronze if you decided to replace the PSU.
Cooler master MWE 500w

I am using msi 1060 3gb with dual fans, assuming yours is reference design blower fan so can't really speak to the typical temps for that reference design aside from citing max temp spec. I also have a mini 1060 with 1 fan that never got near 90c though. I don't see you getting out of that situation without some cash, but maybe someone else can help you diagnose the problem better then my personal suggestions.

 

[Backhander]

Not a problem, I am merely suggesting things I would try personally. I have multiple spare power supplies, video cards (1060s/650s) and a backup board/cpu so I can't really suggest everything I would try as I am sure you are limited to the rig you have only. The symptoms appearing without GPU temps being high is troublesome though. I pulled this from nvidia website:
https://nvidia.custhelp.com/app/ans...maximum-operating-temperature-and-overheating

"If a GPU hits the maximum temperature, the driver will throttle down performance to attempt to bring temperature back underneath the maximum specification. If the GPU temperature continues to increase despite the performance throttling, the GPU will shutdown the system to prevent damage to the graphics card. "

So it will shutdown the system to prevent damage to the card if the temp increases past maximum; that doesn't really answer why you would get a crash without the temp being abnormally high. As mentioned before the PSU could be a contributing factor to your system instability, a rail could be dead but the psu still will turn on. Personally I would go through testing ram just to be sure, substitute out the psu for a better quality psu (testing kits for psu are about 15$ though) and test another gpu as that one is reaching the point of interrupting system stability. I would be starting with the ram first as it would be easiest for me to test (not relating to gpu temps), then for me replacing the gpu would obviously be a requirement as it shouldn't come close to 97c even under full load, especially with proper airflow, but for you, you could always increase gpu fan curve to have the fan at 100% once card hits 65c if you wanted to attempt to see if that makes any significant difference, doubtful it will though, you could lock the clock at a lower frequency too, like 1600/1700ish if you desired to try that.
Locking GPU clock (Youtube)
Setting more aggressive fan curve (Youtube)

I can't help you much with mini dumps (assuming you have actual blue screen or kernel dumps to share after enabling the .dmp files). That is the only reason I would personally go through testing ram with memtest to make sure that was not contributing factor to the freeze/crash without card being high temps, and its free/easy to do and removes that as a factor from the issues leaving the psu/gpu as likely source(s) of the problem. The PSU/GPU issue will most likely cost money to avoid the symptoms, if you come to the reasonable conclusion it is the cause of your crash/freeze. I'm not personally confident the GPU is the cause of all your crashes/freezes, but certainly hitting 97 you would reasonably believe its going to shut down to avoid damage to the card.

PSUs are not terribly expensive either, you don't need like 80 plus platinum rated one for a 1060 n i5-8400 anyway its a lower-mid tier system, so I don't really see any issue with looking into one of the 80 plus bronze if you decided to replace the PSU.
Cooler master MWE 500w

I am using msi 1060 3gb with dual fans, assuming yours is reference design blower fan so can't really speak to the typical temps for that reference design aside from citing max temp spec. I also have a mini 1060 with 1 fan that never got near 90c though. I don't see you getting out of that situation without some cash, but maybe someone else can help you diagnose the problem better then my personal suggestions.

I have replaced the graphics card for an older one and the PSU for a new one, tested the RAM and... I still get crashes...

My guess is its either the motherboard or the ssd then?

 

[AdamG]

I have replaced the graphics card for an older one and the PSU for a new one, tested the RAM and... I still get crashes...

My guess is its either the motherboard or the ssd then?

Memtest gave no errors after a few passes of running, and the timings are correctly matching in bios to the product specs of that ram module? If so I suppose running 1 stick at a time checking if symptoms still occur would be redundant at that point having passed memtest and with correct timings.

Do you have a typical samsung ssd? (i didnt see the drive listed in original specs). You could check drive health in samsung magician utility if you have a samsung ssd. You could also run "sfc /scannow" in a elevated command prompt, which checks for system file corruption, although never experiencing issues with my ssds personally I wouldn't be inclined to tell you the most efficient way to check your ssd for issues and would rather defer to someone with experience when dealing with a potential ssd/motherboard issue.

Its good you were able to remove the other gpu and psu from the equation, as thats a pretty fundamental part of basic troubleshooting with hardware problems, even for novice experienced people who have not done alot of builds(like myself) or no builds themselves. I'd wait for someone else to give you additional advice as to what you could do with further troubleshooting in regards to the SSD/Motherboard. I'll keep checking back though, as I am curious what the conclusion will be.

 

[Backhander]

Memtest gave no errors after a few passes of running, and the timings are correctly matching in bios to the product specs of that ram module? If so I suppose running 1 stick at a time checking if symptoms still occur would be redundant at that point having passed memtest and with correct timings.

Do you have a typical samsung ssd? (i didnt see the drive listed in original specs). You could check drive health in samsung magician utility if you have a samsung ssd. You could also run "sfc /scannow" in a elevated command prompt, which checks for system file corruption, although never experiencing issues with my ssds personally I wouldn't be inclined to tell you the most efficient way to check your ssd for issues and would rather defer to someone with experience when dealing with a potential ssd/motherboard issue.

Its good you were able to remove the other gpu and psu from the equation, as thats a pretty fundamental part of basic troubleshooting with hardware problems, even for novice experienced people who have not done alot of builds(like myself) or no builds themselves. I'd wait for someone else to give you additional advice as to what you could do with further troubleshooting in regards to the SSD/Motherboard. I'll keep checking back though, as I am curious what the conclusion will be.

Memtest gave no errors after a few passes of running, and the timings are correctly matching in bios to the product specs of that ram module?

Yes exactly. I will probably seek some proffessional help soon, but not right now, cuz my financial situation isnt very great.
altho i will update u for what happened. thanks!!