Gaming crashes with PC BSODs and SSD performance hitting maximum levels
Gaming crashes with PC BSODs and SSD performance hitting maximum levels
Over the last couple of weeks I've been attempting to resolve a recurring BSOD issue. All failures happened during game launch, while playing, or even when trying to turn on screen share in Discord. My hardware setup includes:
- Motherboard: Gigabyte AORUS B450 M with latest F67d BIOS CPU
- CPU: Ryzen 7 2700
- RAM: Kingston 16 GB (2x8GB) at 2400MHz
- GPU: Palit RTX 4070 Ti
- Graphics card: Samsung 980 Ti Gamerock
- Storage: SSDs – Samsung 980 500GB M.2 (firmware 3B4QFXO7), another SSD – Crucial BX500 240GB SATA
- HDDs: WD Caviar Blue 1TB (games), Seagate Barracuda 1TB (files)
- Power supply: Zalman ZM700-LEII 700W
- OS: Windows 11 x64
Initially, BSODs appeared only at startup or during gameplay. Some examples:
- *Need for Speed Unbound* – fine start but crashes after 15–30 minutes with CRITICAL_PROCESS_DIED or KERNEL_DATA_INPAGE_ERROR.
- *Girls' Frontline 2:Exilium* – freezing and CRITICAL_PROCESS_DIED error after enabling screen share in Discord.
I used standard tools (DISM, sfc, chkdisk) but kept getting clean results despite persistent crashes. After reinstalling Windows, the issue lingered. I re-installed the OS, tried diagnostics, and even checked cables, stress tests, and component health, yet errors kept appearing during gameplay.
When I moved the game to a different SSD and disabled the M.2 drive, the crashes stopped temporarily. But after rebooting, they returned. I examined RAM sticks individually, disks separately, and found the SSD issue resolved only when I swapped it out. Still, problems resurfaced the next day.
Other clues:
1. System freezes during games even with a full 2TB HDD or 240GB SSD.
2. Disks show issues only on the installed SSD.
3. RAM tests didn’t reveal faults despite repeated checks.
4. S.M.A.R.T. and Samsung software report the SSD is healthy, with no errors logged.
5. Firmware was updated to the latest version (3B4QFXO7), but performance remained unstable.
I suspect a hardware fault or intermittent failure in the M.2 drive, possibly due to power delivery or thermal issues. I’m open to suggestions—anyone have insights on this setup?
It might be a bad drive, like the Samsung 980. Move all your data from the Crucial SSD to another hard drive, then install Windows there. Unplug the 980 completely. You could also test it on another computer to check the results. If you don’t have a spare, ask a friend for help.
According to HWiNFO, CMOS voltage is approximately 3.2 volts, which seems acceptable. Unfortunately, I don’t have a multimeter for direct testing. I could test with a CPU from my cousin’s PC (Ryzen 5 2600x), but I’m missing thermal paste, so swapping CPUs or adjusting the cooler might be necessary. I’m planning to purchase a Ryzen 7 5700X3D with thermal paste around mid-January. Thanks!
It seems my machine has several components that may need updating, since it was assembled a long time ago. Only the M.2 SSD and graphics card are fairly recent additions. Although the power supply is outdated and basic, it still performs adequately according to system readings, as it rarely exceeds half its capacity. I intend to replace it soon—maybe with the CPU or later. Appreciate your feedback!
The battery voltage reading isn't accurate—it changes when the PC is running. Swap in a new battery, it's inexpensive, and your board is old enough. Also, the cousin who has the 2600X might have some paste left over.
The readings aren't very reliable and the speed is too slow to catch any minor power hiccups. The issues stemming from instabilities suggest either a damaged storage drive or a faulty power supply unit that behaves as expected. You'll likely need a defective PSU to trigger problems. Compare these models with similar ones from Thermaltake—some perform just as well or better, using less than a quarter of their maximum output.
Critical_Process_Died often signals various issues, but your logs lack clear clues. If storage is faulty, this error tends to appear, though it doesn’t confirm poor storage quality. A Kernel_Data_Inpage_Error suggests the page file became corrupted, which is a serious warning about storage health. It might also point to memory problems since the page file acts like additional RAM. You can't pinpoint the exact origin, only where the problem was detected. Memory issues aren’t limited to RAM or storage—they can involve the CPU as well. Unfortunately, SMART monitoring isn’t very helpful for NVMe SSDs (and some SATA SSDs with new reporting features). Overall system health remains questionable, depending on how much failure time each manufacturer allows before alerts trigger. The reported percentage reflects how much of the expected write capacity remains. Previously, checking SMART parameters on HDDs and SATA SSDs gave fairly reliable results (about 80% of failed drives showed signs), but with NVMe SSDs, useful data disappeared. Out of many faulty NVMe drives I’ve seen, fewer than five displayed any warning signs in SMART checks. Samsung Magician mainly scans SMART, but it hasn’t helped me identify bad drives consistently.