[SOLVED] Pc crashes down whan playing games

[Riki-Oh]

Since I built a new pc, I have this problem where while playing 3d games or running programs like Premiere Pro, benchmarks, or even for some reason Samsung Magician, the pc crashes and reboot without the bsod.
I think I've tried pretty much everything on the software side, and the hardware is all brand new. The problem doesn't seem to be the CPU, GPU, or NVMe overheating, nor a compatibility issue of some programs.
I have tried to re-install windows, all the programs and all the drivers (I also tried older versions of the GPU driver and windows 10), and I tried to install the operating system in my second SSD and removing the NVMe and vice versa, but the problem remained.
I thought the problem may have been a power supply(I had a 750 watt Corsair RM750), so I replaced it with a 1000 watt Corsair HX1000i, still nothing.
Do someone know what my problem can be?

Pc specs:
OS: Windows 10 64 bit version 2004
CPU: Intel Core i7-9700K
GPU: ASUS ROG STRIX RTX 2080 Super OC edition
PSU: Corsair HX1000i (1000 Watt)
MB: ASUS ROG STRIX z390-e gaming
Nvme (Windows and all programs): Samsung 970 Evo plus 500 GB
Ssd Sata III (Only games): Samsung 860 Evo 250 GB
CPU Cooling: Corsair Hydro 100i RGB Platinum SE
Memory: Corsair Vengeance pro 32 GB 3200 Mhz

 

[Riki-Oh]

Well, does the issue remain when you disable xmp?

No it wouldnt be normal for the mhz to be 1000 when xmp is enabled vs 1067mhz when its disabled..

Can you test a different kit of ram?

I was saying the frequency was about 1000, not exatly 1000. I didn't know that normally the frequency is stable at exatly 1066/1067. And I'm not exatly sure anymore I read right when I cecked the frequency with xmp enabled.

But now I have another, bigger problem. I tried to reinstall windows and now the pc doesn't work at all anymore.

It tries to do an automatic rapair but it gives me an error and I can't do anything else than shut down the system and try again....

I already had another similiar problem before, where the pc crashed every 10 minutes or so and when I tried to reinstall windows it just continued to fail, and I thought I solved it after changing the motherboard, the gpu, the ram and the cpu in this order. I swear I'm going to cry, I tried to solve this <Mod Edit> for one <Mod Edit> year, from when I first built this pc, and now, after changing litteraly every single piece of the hardware and without any weird softwares or viruses, this computer still doesn't work....

What am I doing wrong?

 

[Vic 40]

It tries to do an automatic rapair but it gives me an error and I can't do anything else than shut down the system and try again....

Reinstall or reset?

 

[Riki-Oh]

[

Well, does the issue remain when you disable xmp?

No it wouldnt be normal for the mhz to be 1000 when xmp is enabled vs 1067mhz when its disabled..

Can you test a different kit of ram?

I checked again with XMP on and I was wrong, the ram works at 1600 Mhz with XMP on and 1060 with XMP off.

QUOTE="Vic 40, post: 21922070, member: 1290656"]
Reinstall or reset?
[/QUOTE]
Sorry, I ment reset. But somehow now works, I had to turn the pc on and off a number of times, set the cpu voltage to 1,100, again on and off 5/6 times and then set the cpu voltage on auto and somehow everything worked well again, and widows resetted successfully.

 

[Vic 40]

So ram is working at rated speed.

And no more problems?

 

[Riki-Oh]

So ram is working at rated speed.

And no more problems?

The main problem remains, the pc still crashes when I play games or use benchmarks

 

[Jason H.]

[

I checked again with XMP on and I was wrong, the ram works at 1600 Mhz with XMP on and 1060 with XMP off.

QUOTE="Vic 40, post: 21922070, member: 1290656"]
Reinstall or reset?
Sorry, I ment reset. But somehow now works, I had to turn the pc on and off a number of times, set the cpu voltage to 1,100, again on and off 5/6 times and then set the cpu voltage on auto and somehow everything worked well again, and widows resetted successfully.

I would test another kit to narrow down the issue. If you have been having issues with the ram speed its possible 1 of the chips are going out. If you dont have another kit to test, at least try 1 stick at a time. See if the issue persists or goes away with only 1 of the 2 chips installed at once.

 

[Riki-Oh]

I would test another kit to narrow down the issue. If you have been having issues with the ram speed its possible 1 of the chips are going out. If you dont have another kit to test, at least try 1 stick at a time. See if the issue persists or goes away with only 1 of the 2 chips installed at once.

I don't have a second kit, but I tried using one stick at a time and the problem still remained

 

[SM/Madao]

I don't have a second kit, but I tried using one stick at a time and the problem still remained

can you try Benchmark with Aida64 Extreme? https://www.aida64.com/downloads

Navigate to "Tool" panel and Benchmark GPGPU and also RAM, ignore the Trial message from benchmark.

 

[GrandSACHI]

I am curious to know which benchmark softwares have the computer to crash. That might give an idea of which components lead to the crashes when stressed.

Even Samsung Magician provoke a crash? Could it be your SSD?

 

[Riki-Oh]

I am curious to know which benchmark softwares have the computer to crash. That might give an idea of which components lead to the crashes when stressed.

Even Samsung Magician provoke a crash? Could it be your SSD?

Right, sorry, I forgot to specify that the benchmarks that make my computer crash are graphic/overall performance ones like 3dMark.

I thought about the ssd thing. But I already changed my main ssd and tried to unplug my secondary one, and anything changed.

can you try Benchmark with Aida64 Extreme? https://www.aida64.com/downloads

Navigate to "Tool" panel and Benchmark GPGPU and also RAM, ignore the Trial message from benchmark.

Ok I did it, these are the resoults.

(I don't know how I should post multiple images so I'll just leave the imgur link)

https://imgur.com/a/nib9cdE

 

[GrandSACHI]

Okay that's odd considering you changed pretty much everything. Out of curiosity, do you get a crash when you remove the GPU and use the iGPU instead?
Also, how would you describe the stability of the electricity in your household? Reliable? Unreliable? I wonder if your electric outlet is unable to provide sufficient power to feed you computer when under load... just a wild guess here.

Could also be the CPU though, since it's pretty much the only thing you did not change, unless I missed something.

 

[Riki-Oh]

Okay that's odd considering you changed pretty much everything. Out of curiosity, do you get a crash when you remove the GPU and use the iGPU instead?
Also, how would you describe the stability of the electricity in your household? Reliable? Unreliable? I wonder if your electric outlet is unable to provide sufficient power to feed you computer when under load... just a wild guess here.

Could also be the CPU though, since it's pretty much the only thing you did not change, unless I missed something.

I'm going to try without the graphics card right now.

Update: It actually works without dedicated graphics card! I'm still concerned that the issue may be something power/cooling related instead of the graphics card, but I'll try to change it and see if the problem disappear.

The electricity in my household is stable and reliable. My family owns other 2 computers and they work completly fine all the time, even when used at the same time.

About the CPU, I actualy changed it 3/4 months ago. But before that I changed the motherboard, the ram and the graphics card and at the time the computer was completely unusable, so I don't know if the issue is one of these components.
I already changed the motherboard a second time, so maybe the problem are: RAM, Graphics Card or CPU (or maybe some weird compability issue with my CPU cooler since it connects to the motherboard via USB and it's the only component I didn't change yet, but I tried unplugging it and the computer still crashed).

 

[GrandSACHI]

Because you get no crashes when using the iGPU I would say that this rules out the RAM as the culprit. Could indeed be the GPU, but a PSU problem is certainly not out of question. So could still be PSU although you've already change this one...

Now that you mentioned the CPU cooler, I'm don't recall if you checked the temperatures of both CPU and GPU when you got the crashes. If you didn't, you can use HWInfo to monitor the temps while stressing the computer to see if that might explain the crashes...

 

[Riki-Oh]

Because you get no crashes when using the iGPU I would say that this rules out the RAM as the culprit. Could indeed be the GPU, but a PSU problem is certainly not out of question. So could still be PSU although you've already change this one...

Now that you mentioned the CPU cooler, I'm don't recall if you checked the temperatures of both CPU and GPU when you got the crashes. If you didn't, you can use HWInfo to monitor the temps while stressing the computer to see if that might explain the crashes...

Is actually possible to have two different psus, different models with different cables, and have the exact same issue?

About the temperatures: I checked them with OCCt while doing the power test for 45 minutes, everything remained at average temperatures during the crashes, GPU, motherboard, RAM and drives. The CPU did get at 100° and stayed like that for almost the entire test, but the pc did not crash, and the processor never reached more than 75° in any other situation, including during the crashes

 

[GrandSACHI]

For the PSUs, that would be really unlikely, but not impossible. Still very unlikely...

Now 100°C that IS problematic and should never get to that range with a proper cooler. Although you did not find direct correlation with temps and crashes it might still be at the root of your problem. Anyhow, I would get a better cooler just because those temps are not normal and can definitely be detrimental over time.

How about the temps of other compartments? GPU? PCH? VRM MOSFET?
You can get all of these with HWinfo.

 

[Riki-Oh]

For the PSUs, that would be really unlikely, but not impossible. Still very unlikely...

Now 100°C that IS problematic and should never get to that range with a proper cooler. Although you did not find direct correlation with temps and crashes it might still be at the root of your problem. Anyhow, I would get a better cooler just because those temps are not normal and can definitely be detrimental over time.

How about the temps of other compartments? GPU? PCH? VRM MOSFET?
You can get all of these with HWinfo.

Well, the cpu was running at 100% for 45 minutes, is it really that bad if it stays at 100° with that workload for all that time?

About the ohter parts, the gpu was around 60°, Pch: 43°, vrm mos: 57°

 

[GrandSACHI]

For those parts the temps are just fine, no problems there.

For your CPU, just keep in mind that the Tjunction for the i7-9700K is 100°C. That is the maximum temperature allowed at the processor die. Normally when the temps reach levels near that limit, the CPU throttle down in order stay at safe temperatures. If it can't cope with the increase of temperature that goes over 100°C, then it will just initiate a shutdown to avoid damage... which sounds like what you experienced with your crashes.
Now I don't think you damaged your CPU with those of temps simply because the CPU would not allow it. However, in the long run, having a CPU running regularly at those high temperature might see its lifespan reduced significantly and this is why you should definitely seek a better cooling solution. You don't know, that might even solve your crash problems!

 

[Riki-Oh]

For those parts the temps are just fine, no problems there.

For your CPU, just keep in mind that the Tjunction for the i7-9700K is 100°C. That is the maximum temperature allowed at the processor die. Normally when the temps reach levels near that limit, the CPU throttle down in order stay at safe temperatures. If it can't cope with the increase of temperature that goes over 100°C, then it will just initiate a shutdown to avoid damage... which sounds like what you experienced with your crashes.
Now I don't think you damaged your CPU with those of temps simply because the CPU would not allow it. However, in the long run, having a CPU running regularly at those high temperature might see its lifespan reduced significantly and this is why you should definitely seek a better cooling solution. You don't know, that might even solve your crash problems!

I don't know...like I said, the only time where the CPU gets to 100° is during this specific benchmark, but every other time, even with games or 3DMark, it doesn't go over 68° at worst, and I don't think it would right after starting a game like Getting Over It. Still, I'll probably change the cooler, but first let's see what happens if I change the graphics card.

 

[GrandSACHI]

Indeed if you can get the chance to try with a different GPU, that would be first thing to do! 😀

 

[Vic 40]

Even if only when using occt is 100°C too much. Would deal with that too.

 

[Riki-Oh]

Indeed if you can get the chance to try with a different GPU, that would be first thing to do! 😀

I'm wating for Asus to give me the OK for asking Amazon to get a replacement. For now I tried to use my brother's gtx 1660 and it seems to work just fine. Maybe the problem is still related to the psu or something but for now it seems to be a malfunctioning gpu.

 

[GrandSACHI]

Great. Don't forget to let us know if that fix the problem!

 

[Jason H.]

I'm wating for Asus to give me the OK for asking Amazon to get a replacement. For now I tried to use my brother's gtx 1660 and it seems to work just fine. Maybe the problem is still related to the psu or something but for now it seems to be a malfunctioning gpu.

Use your brothers psu to test it just to get that out of your head. You seem to be pretty adiment about it being the PSU, so if he will let you, use his PSU to test.

TO me it dont look or sound like a psu issue especially when you replace the gpu and it works fine. Also no crashing when using the iGPU.

I mean COULD the psu have an issue handling a higher output? Sure... but that doesnt look like the case here. I would definately return the gpu asap and get a new one. Even if it is the psu, the psu is a cheaper replacement than the gpu if you go past your return period.

Amazon takes returns for anything, for any reason (which is why I will always choose them over other online stores like Newegg). You should be fine in that regard. Is there an option to return or request replacement on the order on your Amazon account?

 

[Riki-Oh]

Use your brothers psu to test it just to get that out of your head. You seem to be pretty adiment about it being the PSU, so if he will let you, use his PSU to test.

TO me it dont look or sound like a psu issue especially when you replace the gpu and it works fine. Also no crashing when using the iGPU.

I mean COULD the psu have an issue handling a higher output? Sure... but that doesnt look like the case here. I would definately return the gpu asap and get a new one. Even if it is the psu, the psu is a cheaper replacement than the gpu if you go past your return period.

Amazon takes returns for anything, for any reason (which is why I will always choose them over other online stores like Newegg). You should be fine in that regard. Is there an option to return or request replacement on the order on your Amazon account?

I'll try to change the psu with my brother's one, but it is a non-modular one so it'll be a pain...

About the gpu, I just asked for a replacement, it should arrive this friday, but since nvidia should announce the new 3000 series next week I'm thinking of selling the new 2080 super in order to buy a 3080 IF the prices are similiar like some rumors claim, but that would mean that I would not be able to test another high end gpu until a month or two from now... I don't know, I'll see what to do in the upcoming days

 

[Riki-Oh]

Update: I mounted the new 2080 super and now everything seems to work perfectly. I tried to run 3DMark and playing 1,5 hours and the computer worked just fine.
So it seems that the issue was a malfunctioning Gpu.

Thank you all for your help!