Your game freezes for a few seconds, drops to single-digit frame rates, then either crashes to the desktop or slaps a blue screen on your monitor with an error like CRITICAL_PROCESS_DIED or STORAGE_DEVICE_DISCONNECTED. Sometimes the NVMe SSD that holds your games simply vanishes from File Explorer until you reboot. Audio keeps looping, your mouse still moves, but the drive light stops responding — and Event Viewer quietly logs warnings from sources like storahci, stornvme, or Ntfs around the exact time each crash happened. If the crashing only happens with games installed on your NVMe drive while other drives run games fine, you are almost certainly looking at a storage-related problem, not a GPU or CPU issue.
Do not ignore this. A drive that drops out mid-session can eventually corrupt save files, shader caches, or the Windows installation itself, and the cost of a replacement is far lower than the cost of lost data. The good news: most NVMe-related game crashes trace back to a handful of well-understood causes — firmware bugs, aggressive power management, an overheating controller, or a half-seated M.2 stick — and the majority of them are fixable in under an hour without buying anything. Work through the fixes below in order, starting with the easiest, and retest your games after each one.
Likely Causes
| Cause | How to confirm | Fix difficulty |
|---|---|---|
| Outdated SSD firmware | Check firmware version in the manufacturer’s tool (Samsung Magician, WD Dashboard, Crucial Storage Executive) and compare with the latest release on their support page | Easy |
| Aggressive Windows power management (Lpm / HIPM+DIPM) | Drive drops offline during light load or idle gaming; Event Viewer shows stornvme warnings with reset codes | Easy |
| Overheating NVMe controller | Drive temperature above 70°C (80°C+ on PCIe 4.0/5.0 models) in CrystalDiskInfo; crashes worse in summer or long sessions | Easy–Moderate |
| SATA-mode/AHCI or NVMe driver mismatch in BIOS | BIOS storage mode set to RAID or RST Optane instead of AHCI/NVMe on a consumer board | Moderate |
| Drive nearly full or failing (reallocated sectors, media errors) | CrystalDiskInfo health status reads “Caution” or “Bad”; SMART attributes 05, C5, or C6 show non-zero raw values; free space under 10% | Moderate–Hard |
| Unstable PCIe link — bad seating, riser cable, or bifurcation settings | GPU-Z–style link checks or HWiNFO showing the SSD negotiating fewer PCIe lanes than rated; drive installed via extension cable | Hard |
Fix 1: Update the SSD Firmware
What causes it
Manufacturers regularly ship firmware updates that patch controller bugs, including ones that cause the drive to drop offline under load — exactly the failure pattern that crashes games. Early PCIe 4.0 and 5.0 drives in particular received multiple stability patches after launch.
How to check
Open your drive maker’s utility — Samsung Magician, WD Dashboard, Crucial Storage Executive, Kingston SSD Manager, or Corsair SSD Toolbox — and read the firmware version shown on the main dashboard. Compare it against the “latest firmware” listed on the product’s support page for your exact model number.
What to do
Run the firmware updater from within the utility. Most updates are non-destructive and finish in a couple of minutes, but close every game and browser first, plug a laptop into wall power, and let the update run to completion without interrupting it. Restart Windows afterward.
How to undo it
You generally cannot roll firmware back. If a new firmware causes problems, contact the manufacturer’s support for a downgrade tool or replacement guidance — but stability patches almost never need reverting.
Fix 2: Stop Windows From Aggressively Idling the Drive
What causes it
Windows can put an NVMe drive into a low-power state during brief idle moments. If the drive is slow to wake — a common bug on certain controllers — the game’s read request times out and the engine crashes or the drive disconnects.
How to check
Press Win + X and choose Device Manager. Expand Storage controllers (or Disk drives), right-click your NVMe device, choose Properties, open the Power Management tab, and see whether “Allow the computer to turn off this device to save power” is checked.
What to do
Uncheck that box and click OK. Then open Control Panel → Power Options → Change plan settings → Change advanced power settings → PCI Express → Link State Power Management and set it to Off for your active plan. On laptops, do this for the “Plugged in” profile at minimum.
How to undo it
Re-check the box in Device Manager and set Link State Power Management back to Moderate power savings if you want to restore battery-oriented defaults.
Fix 3: Cool Down the Drive
What causes it
NVMe controllers throttle or shut down when they overheat. Drives sitting between a hot GPU and CPU, or PCIe 5.0 models without a heatsink, can easily exceed safe temperatures during game loading and shader compilation.
How to check
Install CrystalDiskInfo (free) and watch the temperature readout while playing for 20–30 minutes. Sustained readings above 70°C — or 80°C on PCIe 5.0 drives — indicate a thermal problem. CrystalDiskInfo will flag the health status with a temperature warning icon.
What to do
First, confirm the factory thermal pad or motherboard heatsink is actually installed and screwed down. If the drive is bare, add a heatsink — even a basic $10 finned model helps. If it already has a motherboard shield, remove the drive, clean the old pad residue with isopropyl alcohol, and reattach with the included pad, checking that it makes full contact. Improving case airflow near the M.2 slot (adjusting a front fan or adding an intake) also helps. Do not stack a thick aftermarket heatsink on a drive under a motherboard’s own heatsink unless the board’s spec allows it.
How to undo it
Remove the aftermarket heatsink and reinstall the original motherboard shield, reusing a fresh thermal pad if the original was damaged.
Fix 4: Free Up Space and Trim the Drive
What causes it
NVMe drives slow down dramatically and can stutter or time out when nearly full, because the controller has no spare blocks to work with during writes. Save files and shader caches constantly writing to a packed drive make game stutters and crashes more likely.
How to check
Open Settings → System → Storage and look at the drive’s free space. Anything under roughly 10% free is a problem. Also run Win + R → dfrgui → Optimize Drives and check when the SSD was last trimmed.
What to do
Uninstall or move games until at least 15–20% of the drive is free. Then, in an elevated Command Prompt (right-click Start → Terminal (Admin)), run defrag C: /O (substituting your game drive’s letter) to send a TRIM command, and fsutil behavior set DisableDeleteNotify 0 to confirm TRIM is enabled. Set Optimize Drives to run weekly on a schedule if it isn’t already.
How to undo it
Nothing to undo — free space and scheduled TRIM are the correct defaults. Reinstall games as needed.
Fix 5: Update or Reinstall the NVMe Storage Driver
What causes it
A corrupted or generic Windows stornvme driver, or an old vendor RAID/RST driver left over from a previous motherboard, can mishandle drive resets and cause disconnects under load.
How to check
In Device Manager → Storage controllers, note the driver name and date (right-click → Properties → Driver tab). If you see “Intel RST” or “RAID” on a board where your SSD should be running plain NVMe/AHCI, that’s a red flag. Also check Event Viewer → Windows Logs → System for repeated stornvme Event ID warnings near crash times.
What to do
If a vendor RAID driver is present, install the plain Windows NVMe driver via right-click → Update driver → Browse my computer → Let me pick, choosing “Standard NVM Express Controller.” Otherwise, download the latest chipset driver package from your motherboard maker’s support page for your exact board model and run it. Reboot afterward.
How to undo it
In Device Manager, use Properties → Driver → Roll Back Driver if the option is available, or reinstall the previous chipset package downloaded from the motherboard vendor’s archive.
Fix 6: Reseat the Drive and Inspect the PCIe Link
What causes it
A partially seated M.2 stick, oxidized contacts, a cheap PCIe riser/adapter, or a wrong BIOS bifurcation setting can cause intermittent link drops — the drive disconnects for a fraction of a second and whatever was reading from it crashes.
How to check
Shut down fully, flip off the PSU switch, and inspect the drive: the standoff should be snug and the screw torqued down without the connector end floating. In Windows, open HWiNFO64, expand your drive, and check the current PCIe link width and speed versus the rated spec (e.g., x4 at PCIe 4.0). If it reports x2 or x1, something is wrong at the physical or BIOS level. If the drive sits on a riser or in a PCIe adapter, test it directly in the motherboard M.2 slot.
What to do
Reseat the drive: unscrew it, pull it out, gently clean the gold contacts with a dry microfiber cloth or isopropyl alcohol, and reinsert it at a firm angle before securing the screw. In BIOS, find the PCIe/Chipset menu and confirm the M.2 slot is set to its native x4 mode, not forced bifurcation (x2/x2 or x4/x4). If the drive is in a secondary slot, check your motherboard manual — some M.2 slots share lanes with SATA ports or run at reduced speed.
How to undo it
Restore the original BIOS PCIe setting and reinstall the drive exactly as it was, noting the standoff position you removed it from.
Fix 7: Test the Drive’s Health and File System
What causes it
Failing NAND, a worn controller, or file-system corruption from previous crashes will keep producing errors even after firmware, power, and thermal issues are ruled out. Repeated hard crashes can themselves corrupt the volume.
How to check
Open CrystalDiskInfo and read the health status. “Good” is fine; “Caution” or “Bad” means act now. Check SMART attributes: non-zero values in Reallocated Sectors (05), Pending Sectors (C5), or Uncorrectable Errors (C6) point to hardware failure. Then open an elevated Command Prompt and run chkdsk D: /f (game drive letter) to scan for file-system errors.
What to do
Let chkdsk repair errors and reboot. Back up saves and any irreplaceable data immediately — this matters even if the drive reports “Good.” If SMART shows failing attributes, replace the drive: clone it to a new SSD with a tool like the free clone features in Samsung Magician or Macrium Reflect Free, or reinstall your games fresh and re-download saves from cloud sync (Steam Cloud, Xbox/Game Bar cloud saves, etc.).
How to undo it
chkdsk repairs are permanent by design. If you replaced the drive, you can keep the old one as a secondary for non-critical storage only if SMART was actually clean.
Fix 8: Update BIOS and Chipset Firmware as a Last Software Step
What causes it
Motherboard vendors patch NVMe compatibility bugs — including drives dropping off the bus — in BIOS updates, especially for early PCIe 4.0 and 5.0 platforms. An outdated BIOS on a newer drive is a classic pairing for these crashes.
How to check
Press Win + R, type msinfo32, and note the BIOS Version/Date on the System Summary page. Compare it with the latest version listed for your exact motherboard model on the manufacturer’s support page.
What to do
Download the latest BIOS file from the motherboard vendor, put it on a FAT32 USB stick, and flash it using the board’s built-in utility (usually EZ Flash via the BIOS, accessed by pressing Del or F2 at boot, or a dedicated BIOS FlashBack button). Follow the vendor’s instructions exactly and do not cut power mid-flash. Install the matching chipset driver afterward.
How to undo it
Most boards let you flash an older BIOS version the same way, but downgrading is sometimes blocked. Only update if earlier fixes haven’t resolved the crashing.
If Nothing Works
If the drive still drops out after firmware, drivers, thermals, seating, and BIOS are all addressed, you are likely dealing with genuine hardware failure or an incompatibility between the drive and your board. First, test the drive in another computer or another M.2 slot to isolate which component is at fault — if the crashing follows the drive, replace it; if it follows the slot, the motherboard’s M.2 controller may be failing, and RMA is warranted. Any drive that repeatedly disconnects should be replaced rather than trusted, even if it tests clean: intermittent storage faults get worse, not better. Back up your data before doing anything else, and check the drive manufacturer’s warranty status by serial number — NVMe warranties typically run three to five years, and a disconnecting drive usually qualifies for replacement.
FAQ
Can a bad NVMe SSD cause crashes even when Windows itself runs fine?
Yes. Windows and background tasks touch the drive lightly, while games stream large assets continuously. A drive that only fails under sustained load will feel fine on the desktop but crash games — and Event Viewer’s stornvme entries around crash timestamps are the giveaway.
Should my games be on the same NVMe drive as Windows?
It’s fine either way. A healthy NVMe drive handles OS plus games without trouble. If you’re diagnosing crashes, though, temporarily installing a game on a different drive is a useful test — if it runs clean there, the first drive or its slot is implicated.
Does PCIe 5.0 cause more game crashes than PCIe 4.0?
Early PCIe 5.0 controllers ran hotter and needed more firmware patching, so they saw more disconnect issues at launch. With current firmware and a heatsink, a PCIe 5.0 drive should be as stable as anything else — but if yours keeps crashing, capping it to PCIe 4.0 speed in BIOS is a reasonable diagnostic step.
How do I know if the crash is my SSD and not my GPU or RAM?
Check three things. Event Viewer → System logs: storage sources (stornvme, storahci, disk, Ntfs) pointing at storage; GPU and RAM failures usually log different sources or nothing at all. Crash timing: does it happen at asset-loading moments? And drive visibility: has the game drive ever vanished from File Explorer or required a reboot to reappear? Two or more “yes” answers means start with the SSD fixes above.
