Addressing crashes and instability issues, especially in UE games
Addressing crashes and instability issues, especially in UE games
Hello everyone,
After roughly two weeks of frequent crashes and overall instability on my desktop, I've tried every troubleshooting method available and still haven't found a solution. I'm reaching out for advice from more experienced users. Performance remains consistent across all tasks, and I don't encounter the standard BSOD. The issue seems to stem from an unstable software environment rather than hardware faults.
My PC was assembled in 2021 using an Alder Lake Z690 build. Up until recently, it performed flawlessly—except during a Star Wars: Jedi Survivor session where the game kept crashing repeatedly. I spent considerable time adjusting BIOS settings to stabilize it. I'm uncertain whether these changes or a patch resolved the problem, given the game's generally poor optimization.
Since then, general instability has increased, affecting not just games but everyday usage. The first signs appeared in Firefox with tab crashes, followed by Chrome and other browsers. Even when launching older titles like Halo Wars or XCOM 2, I experienced random shutdowns. Jedi Survivor is now unreliable; it starts without issue but crashes shortly into gameplay, with UE5 identifying an access violation as the cause.
Most games have started failing at random times, and I've also noticed random closures in Halo Wars and XCOM 2. I attempted to restart Jedi Survivor after a successful playthrough, but it now either fails to launch or crashes mid-game, citing the same access violation error.
Hardware-wise, I haven't changed anything for nearly four years. The only modifications were replacing my original DDR5-4800 RAM with G.skill DDR5-6400 units. Since then, I've only adjusted the GPU—since it was out of warranty, I opened up my RTX 3090 in September to replace the thermal pads, which had become excessively hot. I found minor scorch marks on one pad, but no damage to the chips themselves. The GPU is now functioning well, even with better temperature readings.
Software-wise, I've tested numerous BIOS updates, default settings, and disabled features like XMP to see if that helps. Still, no improvement. Stress tests passed without issues, and memory diagnostics (Memtest86) showed clean results after several hours. OCCT tests also confirmed stability.
Despite these efforts, crashes persist in general use. Browsing has become unreliable, and even stress testing has not triggered failures. I've also tried underclocking the CPU and GPU via MSI Afterburner, but with no lasting effect.
The only change that stands out is the replacement of my RAM modules four years ago. The thermal pads were clearly a concern, and addressing them might have made a difference. I haven't touched any other components since then.
I've shared all these details and will continue to provide logs or further information if needed. Thank you for your time and support!
Best regards,
Jashin
There appears to be a gap in the description between those two phrases.
It's tough to describe without visuals, and I should probably retrieve the photos I took while taking apart the card. As I removed the original thermal pads, I spotted one with minor scorch marks—no more than a millimeter or two. Beneath this pad, a tiny component, possibly a bypass or decoupling capacitor, was damaged. Even though the card still functioned normally, I cleaned the scorch marks on the capacitor, swapped out all the thermal pads as planned, and reassembled the chassis. It seems unrelated to the problem, because the instability timeline isn’t connected to my work (the capacitor likely failed before then). Just in case, I added pictures of the disassembly in September 2025.
I completed a full Windows reinstall using a new ISO on a USB drive. This resolved the startup and early-game crashes, as all problematic programs now run smoothly. I’m relieved because most components are no longer covered by warranty except RAM. I don’t know which driver or file was damaged, but the Driver Verifier test in Windows helped identify the issue—it forced a safe boot only. The blown capacitor on the 3090 is concerning; anyone could provide insight on its role and whether it can be repaired.
Was reading through the thread about to suggest a fresh Windows install. Glad I was right but more importantly I'm glad you got your computer fixed. Anyways, to replace that component you're going to need schematics of some kind so you can see what exactly is fried and what you need to replace it with. I'm no electrician and have zero idea what that is, but I have enough knowledge of electronics to know that a) stuff like this is really really hard to fix) and b) sometimes there's components that can break without hurting anything. If you were to get this fixed it would likely be very expensive because of the knowledge and skill required to work with these tiny little surface-mount components. If the card's working fine, I don't see any reason to attempt it yourself either, unless you're a pro you stand a big risk of accidentally killing your card.
The immediate system crash during the DriverVerifier check prompted me to consider starting fresh. I’m relieved I didn’t have to swap out the CPU, GPU, or motherboard—they’re all still covered under warranty. Regarding the capacitor, I won’t attempt a DIY fix; my experience soldering electronics is limited, and the parts are too large for safe handling. Trying this with my current skills would likely damage a working card, which isn’t worth the risk. Even if it doesn’t work, such fixes would probably cost as much as buying a used 3090 chip.