You’re copying a big game install or exporting a video, and mid-transfer the drive just vanishes. File Explorer throws an error, the game crashes, or the whole system freezes for a second before the NVMe pops back into Device Manager like nothing happened. It’s one of the more unnerving storage faults because it’s intermittent. The drive works fine at idle, then drops off the PCIe bus the moment you push real throughput at it. If you’ve been chasing this and coming up empty, you’re not imagining things, and it’s usually fixable without a rebuild.
An NVMe that disappears under load almost always points to one of three things: heat, power state settings, or a drive that’s starting to fail. We researched owner reports, motherboard forums, and drive vendor firmware notes to sort the common causes from the rare ones. Below is the order we’d work through, from the free 30-second checks to the point where you’re shopping for a replacement. If you’re already eyeing an upgrade while you’re in there, our guides to the best 2TB NVMe SSD for 2026 and the best 2TB NVMe SSD for gaming cover drives that hold speed under sustained writes.
First check the obvious (the 30-second check)
Before you tear anything apart, rule out the cheap stuff. Reseat the drive. Power down, pull the little M.2 retention screw, lift the drive out, and press it back into the slot at that shallow angle until it seats flat, then re-torque the screw so the far end sits snug. A drive that’s held down loosely can lose contact under vibration or thermal expansion. That’s a five-minute fix that solves more of these than people expect.
While you’re in the case, check which M.2 slot you’re using. On a lot of boards the second or third M.2 slot shares PCIe lanes with a SATA port or the chipset, and populating it can force the drive into a slower or less stable link. Read the tiny print in your motherboard manual. Also open your drive vendor’s dashboard (Samsung Magician, Crucial Storage Executive, WD Dashboard) and check for a firmware update. A surprising number of “disappearing drive” bugs are known firmware issues the vendor already patched.
One more quick thing before you commit to deeper diagnosis: open Event Viewer and filter the System log for disk and controller errors around the times the drive dropped. You’re looking for event IDs like 153 or storahci warnings. They confirm the drive is genuinely leaving the bus rather than the OS just hiccupping. If the timestamps line up with your heavy transfers, you’ve already narrowed the field to the three causes below. That log is also the receipt you’ll want if the drive turns out to be under warranty.
Cause #1: Thermal throttling and overheating
This is the number-one culprit for a drive that’s stable at idle but drops under load. Gen4 drives push a lot of data, and a controller pulling 7,000-plus MB/s generates real heat. When the controller crosses its thermal limit it first throttles, and if it keeps climbing some drives will drop off the bus entirely as a protection measure. That’s your disappearing act. It shows up during exactly the moments that generate heat: long sequential writes, game shader compilation, video exports, big archive extractions.
The fix is airflow and a heatsink. If your drive is bare, use the motherboard’s included M.2 heatsink or add a low-profile aftermarket one with a fresh thermal pad. Make sure you actually peel the plastic film off the pad, a genuinely common miss. Watch the drive temperature in HWiNFO during a sustained write. Idle in the 30s to low 40s Celsius is normal; if it’s spiking past 80 to 85 and then dropping out, heat is your answer. Don’t bury a hot Gen4 drive under a GPU backplate with zero airflow and expect it to behave.
Worth knowing: most NVMe drives run two thermal sensors, and the controller usually runs hotter than the NAND. So the number that trips protection isn’t always the one your dashboard shows by default. HWiNFO will list both. If the controller sensor is the one redlining, a heatsink that only contacts the NAND chips won’t help much, and you want a pad that bridges the controller too. Laptop owners have it harder here since there’s rarely room for a real heatsink. In a thin chassis, a thermal pad between the drive and the metal backplate is often the only option, and it does help.
Cause #2: PCIe power state and link management settings
If temps look fine, the next suspect is aggressive power management. Windows and your BIOS both have knobs that can put the PCIe link into low-power states, and some NVMe drives handle those transitions poorly. The drive tries to wake from a deep idle state under a sudden burst of I/O, mishandles the handshake, and drops off the bus. People report this as random disappearances that don’t correlate neatly with heat.
Two settings to try. In Windows, open the advanced power plan settings and disable “Link State Power Management” under PCI Express (set it to Off). Then reboot and see if the drops stop. In BIOS, look for ASPM and NVMe power state options; disabling ASPM for the M.2 slot has resolved this for a lot of owners. On AMD X670 and B650 boards especially, a BIOS update to a newer AGESA version has quietly fixed a batch of these link-stability bugs, so update your board firmware if you’re a few versions behind. These are free changes. Try them before you spend a dime.
Cause #3: A drive that’s actually failing
If you’ve reseated the drive, cooled it, updated firmware, and disabled aggressive power states, and it still vanishes under load, you may be looking at a drive that’s on its way out. Worn NAND, a degrading controller, or a bad solder joint can all cause a drive to drop out precisely when it’s stressed. Pull the SMART data with CrystalDiskInfo. Look at the health percentage, the reallocated or pending sector counts, and the total bytes written against the drive’s rated endurance. A drive reporting “Caution” health or throwing correctable error counts that keep climbing is telling you something.
Back up anything important right now, before you do more diagnosis. A drive that disconnects under load can go from intermittent to dead without much warning. If SMART is clean and the drive is newer, keep it and keep working the heat and power angles. If SMART looks rough or the drive has years of heavy writes on it, plan the replacement instead of babying a failing part. One caveat worth stating: SMART isn’t perfect. Some controller-level or connector faults never show up in the health stats, so a drive can read as “Good” and still drop out. If you’ve genuinely exhausted heat and power fixes and the drops continue, treat repeated disconnects as their own failure signal even when SMART stays quiet.
Preventive maintenance
Once you’ve got a stable drive, keep it that way. Leave 10 to 15 percent of the capacity free so the controller has room for garbage collection and wear leveling, which also keeps sustained write speeds from collapsing. Check for firmware updates a couple of times a year. Keep a heatsink on any Gen4 or Gen5 drive that lives near a hot GPU, and make sure case airflow actually moves air across the M.2 area. Watch drive temps during your heaviest workload once in a while so a cooling problem shows up on a graph instead of as a mystery crash.
Keep a real backup too, not just the drive itself. The whole reason a disappearing NVMe is scary is that it can take live data with it. A second drive or a cloud sync of your important folders turns a drive failure into an inconvenience instead of a disaster. It’s the cheapest insurance you’ll buy for a build, and it pays off precisely on the day a drive decides to drop out and not come back.
When it’s not fixable: what to replace
If the SMART data confirms the drive is degrading, or it keeps dropping out after you’ve ruled out heat and power settings, replacing it is the honest fix. A reliable, high-endurance drive is worth the spend when your boot volume or working scratch disk keeps vanishing mid-task. The Samsung 990 PRO 2TB is one we’d point people toward here. It’s a PCIe Gen4 M.2 2280 drive rated for sequential reads up to 7,450 MB/s, built for heavy-duty workstation and gaming loads, and at $389.99 with a 4.2 rating it’s a known-solid controller and NAND pairing rather than a bargain-bin gamble. That reliability matters most in the exact scenario that got you here, sustained load.
Pros
- Rated 7,450 MB/s sequential read pushes close to the PCIe Gen4 x4 ceiling for game loads.
- Rated 6,900 MB/s sequential write helps large file transfers and 4K editing scratch workflows.
- Samsung claims 55 percent random IOPS uplift vs the 980 PRO, benefiting bot and app launch responsiveness.
Cons
- Limited owner feedback available at time of writing, so long-term reliability signal remains thin.
- TBW endurance rating not specified in listing, worth confirming for sustained write workloads.
- No heatsink on this SKU, sustained Gen4 writes may need a motherboard M.2 cooler to avoid throttling.
The Samsung 990 PRO 2TB is a high-end PCIe Gen4 NVMe SSD in the M.2 2280 form factor, rated up to 7,450 MB/s sequential read and 6,900 MB/s sequential write. It targets desktop gamers, PS5 upgraders, and content creators who want Gen4 headroom without stepping up to a Gen5 drive.
The defining pitch is near-ceiling Gen4 bandwidth paired with a claimed 55 percent uplift in random IOPS over the 980 PRO. In practice, drives at this tier typically shave a second or two off open-world level loads and accelerate DirectStorage titles, though real-world gaming deltas over solid Gen4 rivals remain modest.
Gen4 flagships like this tend to run hot under sustained sequential writes, and the bare version without a heatsink relies on motherboard M.2 cooling to hold peak speeds. TBW endurance is not specified in the listing, so buyers running heavy write workloads should confirm the rating from Samsung's own datashet.
Buy this if you want a top-shelf Gen4 drive for a gaming PC, editing rig, or PS5 with a well-cooled M.2 slot. Skip this if you already own a fast Gen4 SD, need Gen5 bandwidth for scratch work, or want a factory heatsink model included in the box.
Sequential throughput: Samsung rates the 990 PRO at up to 7,450 MB/s sequential read and 6,900 MB/s sequential write over PCIe Gen4 x4. That is close to the practical Gen4 x4 ceiling near 7,500 MB/s, so large file copies are limited more by the source disk than by this drive.
Random performance: The listing calls out a 55 percent improvement in random performance versus the 980 PRO. Random 4K IOPS drives OS snappiness, shader cache loads, and small file operations, so the uplift matters more day to day than the sequential headline for most users.
Form factor and platform: The drive uses M.2 2280 single-sided layout and PCIe Gen4 x4, dropping into standard desktop M.2 slots, most Gen4 laptops, and the PS5 storage bay. Capacity spans 1TB to 4TB in the family, with this SKU at 2TB for a balanced game library.
Power and thermals: Samsung claims up to 50 percent better performance per wat versus the 980 PRO, helping in fanless laptop bays and compact ITX builds with limited airflow. Under sustained writes, this bare version benefits from a motherboard M.2 heatsink to avoid thermal throttling on long transfers.
Match the replacement to the job. A 2280 form factor fits nearly every desktop board and most laptops, and Gen4 speed is plenty unless you specifically need Gen5. If you’re replacing a boot drive, our roundup of the best 2TB NVMe SSD for a boot drive weighs the options that stay responsive under everyday multitasking. Clone with the vendor’s migration tool or do a clean Windows install, then confirm the new drive holds up under the same workload that killed the old one before you trust it.
Tools and parts needed
You won’t need much. A Phillips #1 screwdriver for the M.2 retention screw, an M.2 heatsink with a fresh thermal pad if your drive is bare, and free software: HWiNFO for temperatures, CrystalDiskInfo for SMART health, and your drive vendor’s dashboard for firmware. That’s the whole kit for diagnosing this. A replacement NVMe only enters the picture if the SMART data or repeated drops tell you the current drive is done.
A few more questions
Can overheating permanently damage my NVMe drive?
Occasional thermal throttling won’t kill a drive; that’s the controller protecting itself. Repeated extreme heat over months can shorten NAND lifespan, though. The bigger risk is data loss when the drive drops out mid-write and corrupts whatever was being saved. Add a heatsink and airflow so it never gets there, and you’ll sidestep both problems.
Why does my drive only disappear during large file transfers?
Large transfers are the workload that pushes both heat and power draw to their peak. A drive that’s fine reading a web page or launching an app can hit its thermal or link-stability limit the moment you throw a 60GB write at it. That’s why the fault looks intermittent. It only shows up when you stress the exact weakness, which is usually heat or a poorly handled PCIe power state.
Should I move the drive to a different M.2 slot?
It’s worth a try, especially if you’re in a secondary slot that shares lanes with SATA or the chipset. Moving to the primary CPU-connected M.2 slot gives the drive its own full-speed link and often better cooling from the board’s main heatsink. Just check your motherboard manual first so you know which slot is which and what disabling shared ports might do to your other drives.
Write Your Review
No reviews yet. Be the first to share your experience!