Question Brand new PC build and every game i play crashes to desktop ?

[kobr81]

Greetings people!

Several days ago i built a brand new PC and im having very strange problem and im getting desperate.

My PC specs:
CPU: Ryzen 7700x
MB: ASUS ROG B650e-e GAMING WI-FI
RAM: Kingston Fury 32GB 6000mhz EXPO certified
GPU: PowerColor RX 7900 XT 20GB HELLHOUND
STORAGE: Kingston KC 3000 nvme 1TB (WINDOWS 10) and XPG GAMMIX s70 2TB for games
COOLER: Deepcool Zero Dark AK620 air cooler
PSU: ASUS ROG 850W Gold
MONITOR: LG 2K 27GP850

Like i said my problem is very peculiar and by that i mean that every game i started to play is crashing to desktop. I retired my old PC on which i never had any problems and played every game without problems. On this new PC i can do every task normaly, everything is stable except games. I think i tried most of the solutions i could think of and have found on internet, now im getting desperate.

First game i installed was Star Wars: Jedi Survivor and game crashes every 1-2 minutes while playing. I thought ok its a buggy game so i installed Witcher 3 and same thing there it crashes after 3-4 minutes. Now im getting worried and then i installed MechWarrior Online and i managed to play like 20 minutes before it starts to crash to desktop. Fallout 4 also crashes to desktop after few minutes...

I ran multiple stress tests on my system and everything went fine (cinebench, memtest, furmark, heaven unigine ...) no crashes whatsoever. I did this with EXPO enabled, disabled its all the same games crash and stress test is fine. BIOS is up to date, gpu drivers also, chipset is up to date. I even reinstalled windows from scratch and its same all over again.

One thing i noticed is in Jedi Survivor and in that game while on main screen cursor becomes pixelated like it has artifacts and i can move it around like that, i saw this pixelated cursor in MechWarrior Online on 2 occasions. I have image of this how it looks like:

https://imgur.com/ePHZlGY

View: https://i.imgur.com/ePHZlGY.jpg


I suspect its GPU at fault here but would like your expert opinion on what else i could do. I paid good money for this PC, everything in my country is so expensive and that this happens.

EDIT: Forgot to mention that i disabled all overlays, from steam, afterburner... and it doesnt help even when i close every other app i have. I did not se high temparatures, CPU is around 60c in games and GPU also around 60c.

EDIT 2: Also i noticed that my screen would flicker when i open my web browser (EDGE), and my rgb on keyboard sometimes would flicker when i open apps but its barely noticable.

Thanks in advance!

 

[Ralston18]

Look in Reliability History/Monitor and Event Viewer.

Either one or both tools may be capturing some errors codes, warnings, or even informational events just before or at the time of the crashes.

Could be heat related. Works, gets hot, crashes, cools down/restart, heats up again. Loop.

Thermal paste: what was used and how was the paste applied?

= = = =

As a general observation and suggestion:

Power down, unplug, open the case.

Verify by sight and feel that all connectors, cards, RAM, jumpers, and case connections are fully and firmly in place.

No one wants to force anything especially with a new PC. However what feels to be tight and in place may not really be so.

Carefully unseat and reseat everything a few times. You may discover something that suddenly feels smoother and more "solid" as it goes into place.

Take your time. Use a bright flashlight to inspect everything as you go along.

 

[kobr81]

Look in Reliability History/Monitor and Event Viewer.

Hello and thank you for your response. I found something in reliability monitor when Witcher 3 crashed and it says this:
Problem signature
Problem Event Name: APPCRASH
Application Name: witcher3.exe
Application Version: 4.0.1.22341
Application Timestamp: 00000000
Fault Module Name: witcher3.exe
Fault Module Version: 4.0.1.22341
Fault Module Timestamp: 00000000
Exception Code: c0000005
Exception Offset: 000000000183c840
OS Version: 10.0.19045.2.0.0.256.48
Locale ID: 1033
Additional Information 1: 7f71
Additional Information 2: 7f719da74db03a7ca3aa737943c65505
Additional Information 3: 123c
Additional Information 4: 123c074af7879ee38b3e9f42bebcd2c9

Sadly dont know what it means.

Could be heat related. Works, gets hot, crashes, cools down/restart, heats up again. Loop.

I did not notice any overheating in afterburner and hwinfo. CPU and GPU hower around 60c.

Thermal paste: what was used and how was the paste applied?

I used Arctic MX 4 paste nice evenly spread in x pattern.

Verify by sight and feel that all connectors, cards, RAM, jumpers, and case connections are fully and firmly in place.

No one wants to force anything especially with a new PC. However what feels to be tight and in place may not really be so.

Carefully unseat and reseat everything a few times. You may discover something that suddenly feels smoother and more "solid" as it goes into place.

I reseated my GPU today. Everything else seems fine and in place as it should be. Fortunatly this is not my first PC build so i have some expirience but i will try this if everything else fails.

Is there maybe some other tool that i can test my PC that i maybe missed for RAM or GPU?

Also i edited my original post with image that might be interesting and with some new info.

Thank You!

 

[Ralston18]

Reinstall Witcher3

Also could be some corrupted files.

Run "dism" and "sfc /scannow".

FYI:

https://www.windowscentral.com/how-use-dism-command-line-utility-repair-windows-10-image

https://www.lifewire.com/how-to-use-sfc-scannow-to-repair-windows-system-files-2626161

 

[kobr81]

Run "dism" and "sfc /scannow".

Hello again. I ran sfc/scannow and it found some errors and repaired them. Now my cursor in Jedi Survivor looks normal. But the game still crashes.

Today for the first time when Jedi Survivor crashed i got BSOD and before it dissapeared i managed to read in bottom left corner ATTEMPTED TO WRITE TO RANDOM ONLY MEMORY or something like that. Luckily i saw in another thread where someone has similar problem of an app called OCCT for stress testing.

So i ran OCCT and run RAM test and my God only after few minutes i got over 6000 errors, program was stuttering heavily and i barely managed to stop the test. After that i tested VRAM, CPU and Power in that same app and everything passed tests except RAM.

I will see today if RAM sticks are properly in place but i doubt it will help.

I will update if i solve the issue with that.

If anyone has some other ideas please dont hesitate to share them

Thank you!

 

[Ralston18]

Memory:

"MB: ASUS ROG B650e-e GAMING WI-FI
RAM: Kingston Fury 32GB 6000mhz EXPO certified"

RAM =Single 32 GB? Source?

Perhaps= 4 x 8GB or 2 x 16GB? Matched set/kit for dual channed use?

Refer to the motherboard's User Guide /Manual to ensure that the installed RAM is both supported and installed in a supported configuration.

Also visit the motherboard's manufacturer's website for more up-to-information, changes, or corrections regarding RAM.

Some motherboards require that the first physically installed RAM module be placed in a specific slot.

 

[kobr81]

RAM =Single 32 GB? Source?

Perhaps= 4 x 8GB or 2 x 16GB? Matched set/kit for dual channed use?

Sorry i was supposed to write 2*16 GB RAM kit, mathed kit for dual channel use sloted in a2 b2 slots as per motherboard manual.

 

[saperos]

@kobr81 Hey, I wanted to mention that I was able to resolve similar problem by reducing the max PBO frequency boost. Please find details here: https://forums.tomshardware.com/thr...desktop-on-my-new-build.3817989/post-23087217

 

[kobr81]

@kobr81 Hey, I wanted to mention that I was able to resolve similar problem by reducing the max PBO frequency boost. Please find details here: https://forums.tomshardware.com/thr...desktop-on-my-new-build.3817989/post-23087217

Interesting. Did you manage to have a normal session playing demanding games with no crashes? I did notice that my CPU boost up to 5.5GHZ on single or couple of cores when playing games.

I have been busy testing things like trying single RAM stick and games still crash even when i use only single stick of RAM. What i noticed is that on one stick i can play like 30 mins before it crashes, and on other stick it crashes like before, every 2-3 minutes depending on a game.

I also reseated my CPU and that didnt help either. I put back everything like it was before and ran OCCT and had a lot of error reports. I also ran TestMem5 and that app also showed RAM errors.

I brought back my old PC and will try swaping GPUs, on my old one i have RX 580 4GB and will check if that makes a difference.

Im almost convinced 80% that its the RAM fault, and 20% on Motherboard, but we will see. Reason i say this is because i noticed in Reliability Monitor when Witcher 3 crashed this error: Exception Code: c0000005. Googled this code and it says its RAM problem.

Worst thing is because this process is so time consuming and no matter what i do cant seem to fix it . At least i hope you solved all your problems.

 

[Biocc]

I had a similar issue with games crashing on a brand new PC and also started out with Jedi Survivor. Started playing Deus Ex Mankind Divided and also got crashes there from time to time. Even viewer was logging this error during my crashes:

dxgi_error_device_hung

What seems to have solved it for me was changing PCIe Gen in BIOS from AUTO to Gen4 or Gen3. Also, Jedi Survivor continued to crash after this, though not as often.

 

[saperos]

Interesting. Did you manage to have a normal session playing demanding games with no crashes? I did notice that my CPU boost up to 5.5GHZ on single or couple of cores when playing games.

I have been busy testing things like trying single RAM stick and games still crash even when i use only single stick of RAM. What i noticed is that on one stick i can play like 30 mins before it crashes, and on other stick it crashes like before, every 2-3 minutes depending on a game.

I also reseated my CPU and that didnt help either. I put back everything like it was before and ran OCCT and had a lot of error reports. I also ran TestMem5 and that app also showed RAM errors.

I brought back my old PC and will try swaping GPUs, on my old one i have RX 580 4GB and will check if that makes a difference.

Im almost convinced 80% that its the RAM fault, and 20% on Motherboard, but we will see. Reason i say this is because i noticed in Reliability Monitor when Witcher 3 crashed this error: Exception Code: c0000005. Googled this code and it says its RAM problem.

Worst thing is because this process is so time consuming and no matter what i do cant seem to fix it . At least i hope you solved all your problems.

Yes, I was able to play for 2 hours straight yesterday! I went through all the steps you mentioned, such as swapping the GPU, PSU, and even buying a new set of RAM to test. However, what ultimately resolved the issue was adjusting the PBO max frequency.

While observing the frequencies in OCCT, I noticed that in tests like 3D Mark and Cinebench, the frequency was staying below 5400 MHz, which is why it never failed. But in games and during the OCCT memory test with Heaven benchmark running in the background, some cores were reaching 5430 MHz, and that's when the errors started flooding the screen.

I highly recommend giving it a try. It's a simple adjustment in the BIOS. You can even try turning off PBO completely and test your CPU on stock settings. You might have been even more unlucky with the silicon lottery than I was.

By the way - when I started researching I found that this problem is not limited to you and me, there is something wrong with either 7000 series, or AGESA codes...

 

[kobr81]

Yes, I was able to play for 2 hours straight yesterday! I went through all the steps you mentioned, such as swapping the GPU, PSU, and even buying a new set of RAM to test. However, what ultimately resolved the issue was adjusting the PBO max frequency.

While observing the frequencies in OCCT, I noticed that in tests like 3D Mark and Cinebench, the frequency was staying below 5400 MHz, which is why it never failed. But in games and during the OCCT memory test with Heaven benchmark running in the background, some cores were reaching 5430 MHz, and that's when the errors started flooding the screen.

I highly recommend giving it a try. It's a simple adjustment in the BIOS. You can even try turning off PBO completely and test your CPU on stock settings. You might have been even more unlucky with the silicon lottery than I was.

By the way - when I started researching I found that this problem is not limited to you and me, there is something wrong with either 7000 series, or AGESA codes...

Dude i think you just solved my problem. I set Max CPU Boost Clock Override to Negative 200 in BIOS and now my CPU is maxing to 5.35GHZ instead of 5.55GHZ. OCCT didnt show any errors and for the first time i managed to play a game over an 1 hour time. True test will be on weekend when i have more time toplay but i think this is it, considering my games crashed every 2-3 minutes.

Thank you so much for this, you saved me a lot of headache!

 

[saperos]

Great news! I am really glad someone benefited from all the pain I had to go through I'll continue testing like crazy with Starfield tomorrow

By the way, according to the specs, the Ryzen 7600X's max boost speed is 5300 MHz (for the 7700x you have it is 5400 Mhz). This is likely the number that AMD sends as part of AGESA codes to motherboard manufacturers.

However, as a really helpful guy from ASRock support explained to me, they conduct additional tests on a range of CPUs and set the boost speed higher than the spec. They choose an average from all the tested CPUs. That's why I saw 5450 on my ASRock motherboard, and you 5550 on your Asus motherboard.

This approach provides better performance on a given Mobo for the majority of people, but makes life miserable for users like us who have lost the silicon lottery.

That said, setting -150 should give you max boost as per the AMD spec for your CPU.