The games are malfunctioning and I have no fresh thoughts.
The games are malfunctioning and I have no fresh thoughts.
R9 5900x 4070ti 32GB 3200Mhz B450 AORUS ELITE 750W CORSAIR. I just updated Windows and upgraded my case for improved airflow. Since then, gaming has become much harder. My main problem is with Cyberpunk 2077—it ran smoothly before the update but now crashes after about 10 to 15 minutes. Marvel Rivals behaves similarly; it closes or freezes completely after short play sessions, sometimes forcing a reboot. I've checked everything online, tried different disks, disabled mods, enabled compatibility, and even ran games as admin. Windows files are intact in CMD, but the issue persists. My CPU stays around 4.4Ghz during gameplay, hovering in the low 30s to high 50s. Anyone have any suggestions?
Crashing as in the game just turns off. For Cyberpunk there is a little box that says "Cyberpunk has Flatlined" and asks you if you want to send the report or not, For Marvel Rivals its an Unreal Engine error or just straight up freezing the pc to the point that you can't do anything else but to turn it off and back on again. All drivers are up to date , BIOS as well , GPU drivers don't seem to be the issue here , rolled back and still have the same issue.
The Nvidia 572.16 version wasn't ready yet. Many users reported problems and BSODs. Some claim issues persist even after updating to an older driver, possibly related to your case.
After the crash, return to your Windows desktop and review the Event Viewer. Navigate to Windows Logs, then System, and scroll through the logs until you reach the timestamp of the incident. Look for any unusual entries—especially those marked as critical—but don’t overlook the many amber warnings that might be normal. Focus on graphics drivers; consider reinstalling them or updating to the latest versions first, keeping older ones ready for comparison. Clear all temporary files and programs, including XMP/EXPO, before testing each update one by one. If everything works after the initial changes, gradually enable features again to identify the problematic component. Right now, you’re running version 566.36 on 566.36 and are not experiencing issues.
Launching Cyberpunk after resetting BIOS settings back to defaults, excluding any overclock or XMP adjustments, brought it to version 566.36 via DDU as recommended. Despite this, the game continued to crash. In the event viewer, the only visible issue was that the application-specific permission settings denied Local Activation for the COM Server application with the specified CLSID and APPID, assigning the user DESKTOP-randomnumbers\user from the localhost address through an LRPC running inside the app container. The system reported an unavailable SID, indicating a security restriction that can be adjusted via the Component Services admin tool. The desktop name was changed, but the user account remains unchanged as shown in the event viewer.
Was active, switched it off for testing as advised by another user.
The situation before the crash seems to have been handled well by Nvidia. Their drivers are built to handle crashes and recover, even during gameplay. In practice, you often encounter a black screen because the driver fails, but audio remains and the system stays operational. The two examples you mentioned fall within the expected normal behavior. As shown in my screenshot, there are many similar cases.
Are you using Windows 10 or 11? Although I don’t hate Windows 11, perhaps a clean installation of Windows 10 could help identify an OS-related issue. Right now it doesn’t really offer much long-term benefit, but I’m running low on software fixes which leaves room for a hardware problem to be overlooked. When you changed the case, did you remove the entire system, including the CPU and cooler? It might be worth setting up a simple bench build using cardboard on your kitchen table. You’d be surprised how often a small misplaced screw or washer behind the motherboard can cause such issues. Or even just a loose power supply connection. Could be that something got damaged during the swap. Re-seat the CPU and cooler. A bad pin-to-pad contact could be the culprit. Try taking everything apart, unplugging, and reassembling while checking for clean contacts and debris. Plug the GPU into a different PCIe slot—this isn’t a permanent fix but it’s worth trying. Add an extra RAM stick and test in another slot. Do you have another PC available for testing? If one of your components is faulty—GPU, CPU, RAM, or motherboard—having another machine to swap parts in and out always proves useful. Stay updated and good luck! This sounds like a real headache!