Issues with games closing to the desktop without any warning signs?
Issues with games closing to the desktop without any warning signs?
Hey guys, first post here! My friend assembled new parts (CPU, RAM, MOBO, PSU) to my PC in late November. After few weeks of gaming I started to encounter weird crashes to desktop. No bluescreens etc. Desktop is running smoothly, and all the problems are occuring while ingame. Doesn't matter which game I play, everything shuts down suddenly. For example CS2, Lethal company, WoW and Rocket league. Weird part is that they typically do not leave any error codes that could help me troubleshoot. Nothing relevant to event monitoring nor system information either (atleast I cant make any use of it with my knowledge).
SPECS:
GPU: Nvidia 3080 Gaming Z Trio (OLD)
CPU: Intel i9-14900k (NEW)
RAM: Corsair 2x 16gb 6200mhz (CMH32GX5M2B6200C36) (NEW)
MOBO: ASUS TUF-GAMING-Z790-PLUS-WIFI (NEW)
PSU: Corsair 1000W RM1000x (2021) modular ATX-PSU (NEW)
AIR COOLING: Noctua NH-15D (OLD)
Few HDDs and SSDs (OLD)
Things that I have done already, but crashes keep occuring:
Reinstalled games to different SSD/HDD
Updated BIOS to latest one
Rolled back to previous GPU driver (didnt work so switched back to most recent update)
Updated every driver I could ever find to hardware
Updated to Windows 11
Reinstalled windows from desktop, and from bios with USB
Ran Memtest86, 0 errors
Ran windows ram test (mdsched.exe), 0 errors
Bought new RAM's
No OC on GPU or CPU
Thermalpaste reapplied, temperatures shouldnt be a problem
Ran windows repair tool sfc/scannow
Ran 20mins of Furmark, nothing unusual
Ran few antimalware scans, nothing unusual
Ran "
Intel® Processor Diagnostic Tool
", CPU passing every "test"
Only game that has ever given me error code is WoW. "Program: H:\Battle.net\World of Warcraft\_retail_\Wow.exeException: ACCES_VIOLATION - The instruction at "0x00007ff6970e9bdf" referenced memory at "0xffffffffffffffff";The memory could not be "read",ProcessID: 31436ThreadID: 40044"
Obviously this being the only hint I reached on to Blizzard support for more detailed information and they responded me like so:
"Sounds like you were experiencing crashes in WoW and other games. I definitely know that it is never fun when you run into issues like this, and I want to let you know that we will do everything in our power to help you get it resolved.
Looking over the crash and your system files this appears to be a crash related to your NVIDIA GeForce RTX 3080 card. In the Windows error logs we can see many LiveKernelEvent code 141 errors and these would indicate GPU driver timeouts, which means that the GPU crashed or stopped responding for a short time. This could be an issue with the GPU clock speed being unstable or it could be some kind of hardware issue or power delivery issue to the GPU.”
After this incident I immediately switched to my old GPU (1070), and as you might guess, games are still crashing. One weird thing I discovered few weeks ago, I was curious to see what would happen if I tried one of the OC profiles from bios. So I applied the "Fast tuning" preset from bios, and launched to windows. After a while of gaming I noticed that the game crashes have decreased significantly. For example, before the OC preset, I would crash like 20 times a day in WoW but now with OC preset I crash max once per day. With the OC preset I also noticed that in CS2 and Rocket League 90% of the crashes happen right as I'm about to join a match/server. Then after I manage to get in, I might play for 2 hours without crash. Then again, while joining match the game suddenly crashes to desktop.
So in my eyes, the GPU and RAMs should be clear. Temperatures shouldn't be a problem. I think OS stuff is also ruled out, since I've had these experiences with Windows 10, Windows 11 and even after reinstalling the os. So my thoughts are that the CPU/MOBO/PSU is causing the issue. This chain of events has lasted from December to this day. I have not had the energy on this topic after realizing the OC preset kinda helped me a bit. Now I might try something new. What kind of things would you recommend? I'll probably test out with my old PSU (750W) to rule out any PSU related issues in the upcoming days. Might need to test with old MOBO and CPU combo aswell. I have warranty for all of these hardware, but Im in a difficult situation in my life where I need my PC daily to do my school/work stuff so I cant really send the hardware back to the store for at least few months. So tips are welcome on how to locate the root cause of all this mess.
DISCLAIMER: I am fairly bad at tech stuff, my PC assembling experience is limited to changing GPU and RAM. So monkey explanation is sometimes needed for me to understand.
I was running into a very similar issue as you my friend. I went through 72 hours of troubleshooting, trying everything known to man. The issue appears on 13th gen intel with mainly ASUS boards I've noticed. The solution I've found was to change the SVID behavior in the BIOS from "Auto" to something defined. The way to do that is:
Boot to BIOS
Navigate to "AI Tweaker" at the top
Go to "SVID Behavior." It should be right under XMP, Asus Multicore Enhancement, and Intel Adaptive Boost.
Change it from "Auto," and test each setting. In my case, I got stability on "Worst-case Scenario", but still crash in Helldivers 2 so I ran it on "Intel Failsafe" for a bit.
It should run flawlessly on Intel Failsafe, you will just see increased Temps, and increased voltage draw. Let me know if this helps!
Thanks for your response. Yes, I meant exactly that nothing appears on the screen. I attempted to search for reliability history and event viewer before, but as I mentioned, I'm not very familiar with technology, so I'm unsure which parts to look for from the errors. I can share a link to one of the crash results here (sorry, part of the text is in Finnish):
https://gyazo(dot)com/3b3cd5dc49386753375f94c9f684339a
I'll check the "How to use Windows 10 Event Viewer" guide later. It looks like many crashes mention the "nvgpucomp64.dll" file somewhere in the details. I'll post a few more error links here in case they help someone understand what we're dealing with.
https://gyazo(dot)com/70e738ee55fbc609a4fd4a1a88af48da
https://gyazo(dot)com/5ae4d0c1a945d4e7178b45e5bb03beaa
https://gyazo(dot)com/1dad16f2feca543044888d697ed0c42c
Unfortunately, the website didn't let me share links that function properly. I don't know how to navigate this forum.
Hey, thank you for these tips. Will test these out later. So lets say my PC works normally after I set these settings, is my CPU faulty then? Or is this a "feature" of the i9's lol. Meaning by, if I were to hypothetically send the hardware to manufacturer for warranty service and they would give me new product, would I still encounter the same issue with the new CPU (with all the other components I have)?
Intel Failsafe is the standard SVID setting. Unless you adjust it, it usually lowers your CPU's voltage to reduce temperatures. If this doesn't work at default settings, but functions normally, it seems like a random outcome. When it fails with Intel's default configuration, it indicates a faulty processor, and I would proceed with an RMA. Be prepared for increased temperatures with higher voltage and adjust performance expectations accordingly.
I attempted the Intel Failsafe and disabled the OC preset. I played a full CS2 match without any crashes, which hasn't occurred in several months. I was optimistic these settings would help. Later, after logging into WoW, I played for a while but crashed to desktop again. The same happened with Rocket League, it crashed after a short session. What does this mean? Is the CPU faulty? Here are the crash logs from Event Viewer:
Tried these configurations and it kept failing—probably the hardware isn't working properly.
If you can test another processor, I would do it just to verify.
This inquiry seems unusual, but could you clarify what power mode you're referring to in the control panel?