Experiencing three months of frequent crashes while playing games
Experiencing three months of frequent crashes while playing games
It has been more than three months since this problem started, and I've tried numerous solutions. Generally, I can play for about 20 minutes before the game ends abruptly without any error messages—such as pressing alt+F4—but sometimes it lasts a couple of hours depending on the game. For example, with Diablo 4 it crashes every 10 to 25 minutes, whereas Apex plays for over 30 minutes most of the time. This pattern is consistent across all games I play, except for Enshrouded, which ran for six hours yesterday and only crashed twice.
Details:
https://www.userbenchmark.com/UserRun/67210202
The problem began around when I upgraded my operating system to Windows 11, though even reverting back to Windows 10 still caused issues. Initially, I experienced blue screens, but those were due to incompatible RAM on my motherboard and incorrect slots. Eventually, the crashes became constant. There were reports of ICUE updates damaging RAM (around the time these issues started), so I replaced the RAM as well. Could it be related to my motherboard or power supply unit?
What I've attempted:
- Running games on Windows 10/11, changing RAM, and clearing hard drives
- Performing numerous RAM tests, ensuring temperatures stay low while gaming (using lowest graphical settings)
- Keeping Windows updated, installing different graphics drivers, reinstalling C++, adjusting DRM settings
- Playing at various resolutions, using Dx11 instead of 12, launching in Admin mode, enabling/disabling Fullscreen Optimization
- Running games with minimal background apps (Discord, Spotify & Medal clips are the only ones)
- Clearing temporary files and reinstalling games
- Adjusting Nvidia 3D settings, disabling firewall, running antivirus scans
Regarding network stability: it's reliable, with a consistent connection speed of 930MBps upload and 35MBps download.
Important notes:
- I haven’t overclocked the PC or changed any components in the past few years.
- The system remains clean, and I avoid malicious websites or downloads.
- Someone mentioned that an ASUS Rog Strix B450-F might have network port issues; I purchased a separate network port and it didn’t help. A friend suggested it could be related to an M.2 SSD, which caused problems a few years ago due to wear.
Any suggestions or insights would be greatly appreciated.
are you keeping track of temperatures? how much space is your ssd using? you referred to both versions.
1. download and install crystaldisk info, then verify any errors or crashes on the ssd (i’ve never heard of the spcc brand for m2 ssd)?
2. have you looked for BIOS updates? is windows fully updated?
3. your ram is running at base speed 2666mhz. you should open bios and turn on xmp or set the speed manually to 3600mhz. sometimes this requires repeatedly adjusting voltage to match the specs on the ram—i had to raise my ram’s voltage from stock under-powering (1.2v) to the required 1.35v.
once that’s done, you can enable virtualization on the cpu and, if desired, a d.o.c.p. overclock matching the ram speed. i currently have a crucial ballistix 3600 ram running at full speed, with manual overclocking on my 5600x cpu to 4.6ghz.
4. cleandisk and defrag (yes, even ssds can benefit from a defrag or "trim" occasionally).
5. run a memory test on each ram module individually—leave one stick in slot b2, replace it with the other, and test. if both pass, swap them back to A2.
to perform this, you’ll need to enter bios/uefi settings via a key (like f2, f10, or del) during boot. once there, look for an option to run a memory diagnostic.
your motherboard is a pcie 3.0 board, and the graphics card is a 4.0 model—there might be a slight speed bottleneck.
have you used nvidia ddr to completely remove vid card applications and drivers, and then installed fresh software?
Your 3600 speed ram is running at stock 2133, which could affect performance but won’t lead to crashes. Would you like a detailed list of your components? Specifically, is your 3600 speed ram part of a single matched kit or does it consist of two identical units with different numbers? Have you completed a full memtest86 test without any errors? There should be an option in the ASUS advanced settings to verify this. A failure after some time seems related to overheating, though it may only cause throttling and temporary performance drops—running continuously won’t worsen it. Could you tell me the make/model of your PSU? How long has it been installed? Have you connected the 8 pin CPU EPS connector correctly on the motherboard? Nvidia cards from the 3000 series sometimes experience brief spikes in power consumption that some PSUs might struggle to manage. Also, try running the system without Discord and other background applications; a crash could result if one of them fails. The current BIOS version is 5302: ROG STRIX B450-F GAMING | Motherboards | ROG Global. It improves system stability. You can check your CPU-Z or BIOS settings for details. This might be the first step to resolving the issue.
Changed RAM to 3600 and updated BIOS for that!
OS: Windows 11 Pro 64-bit (switched from WIN10 recently)
CPU: AMD Ryzen 5 5600X (3-4 years old)
RAM: Corsair Vengeance LPX DDR4 3600 C18 2x16GB (same matched kit, week old)
MB: ROG STRIX B450-F GAMING (AM4) (4-5 years old)
GPU: Nvidia RTX 3070 (2-3 years old)
Drives© SPCC M.2 PCIe SSD (red) (4 years old)
(D) Seagate ST2000DM008-2FR102 (2 years old)
Soundcard: Sound BlasterX AE-5 Plus (4 years old)
CPU cooler: Corsair H100i RGB PLATINUM AIO Liquid CPU (5 years old)
Network card: TP-link 2.5 GB PCIe network adapter (couple months old)
PSU: EVGA 700 B1, 80+ BRONZE 700W (100-B1-0700) (4-5 Years old)
Completed several Memtest86 tests and everything was clean each time; currently working on individual drives. The 8-pin CPU eps connector is connected too. Even when running a game alone, it crashed around the 15-minute mark.
Completed all the steps and am currently handling individual Ram checks.
Drive C has 244GB available out of 953GB.
Drive D has 1.7TB free space from a 1.8TB drive.
No issues were detected during sfc or dism/cleanup processes.
CrystalDiskinfo link: http://www.mediafire.com/folder/vvyc784m...95j/shared
3000 series graphics cards sometimes experience unusual spikes that older power supplies can't manage. The Evga 700w model only comes with a 3-year warranty, which doesn't reflect strong confidence in its build quality. Consider testing it with a more robust PSU offering at least a 7-year warranty. After about five years, components like the pump may fail or clog, or air could enter the system, requiring replacement. Keep an eye on CPU temperatures to detect any overheating spikes above 85°C. Sound cards often cause problems and tend to interact with the CPU and motherboard in unexpected ways. You might want to reinstall drivers or test the card without it attached.