Corrupted SSD: What Actually Happens, and What’s Recoverable

This is one of the most persistent misunderstandings about solid-state drives, and it costs people files. Not because the data is actually gone, but because they assume it is, and either give up on a recoverable drive or try something destructive because “it’s dead anyway.” A drive that won’t boot, won’t show up, or throws corruption errors has not necessarily lost anything. It has lost the ability to tell you where things are. It is one of several common myths about solid-state drives worth unlearning before you touch anything.

There is no reliable published figure for how often each failure type occurs, and we won’t invent one. What is well documented is the mechanism. An SSD keeps a constantly updated translation table between the logical addresses your operating system uses and the physical flash blocks where data actually sits. That table lives in the controller’s own memory and is periodically written back to the drive. Damage the table, or the controller that maintains it, and the drive can be electrically fine while looking completely dead.

The practical version: treat every corrupted, read-only, or non-detecting SSD as a “probably still there” problem until an evaluation says otherwise, not a “definitely gone” one. What happens in the next few minutes decides which one it turns out to be.

When an SSD controller detects that it can’t reliably commit new writes, because of NAND wear past its spare capacity, a firmware fault, or a partially failed write, a well-designed controller locks the drive to read-only rather than risk silently corrupting whatever is still readable. That behaviour is documented in SSD controller design literature and even patented directly (a read-only mode triggered specifically after a failed power-loss-protection backup is one real-world example). The lock is the safety net, not the failure. It is the same protective instinct behind why deleted files disappear for good on these drives, just working in your favor this time.

The most common advice for a read-only drive is to clear it: run diskpart and attributes disk clear readonly, or flip a WriteProtect registry key. Both of those commands exist for drives with a software write-lock, a policy setting, not a failing drive. Run against a drive the controller itself has locked, they force exactly the write the firmware was trying to prevent. If the drive was protecting a damaged map, that write can finish the job of losing it.

The safer order: while the drive is still readable, copy everything you can onto different media first. Only once your files are safely elsewhere does it make sense to try anything that writes to the original drive. By then, whether it clears or not stops mattering.

A RAW SSD is one of the clearest examples of the map-versus-territory problem this whole page is about. The operating system reads a small amount of structural information first, partition type, file system signature, volume metadata, before it can read anything else. If that structural layer is damaged but the actual file data behind it is intact, the drive reports RAW: present, but unreadable, because the map that says where things are is what’s broken.

Here is the part most guides skip: a common first step for a RAW drive is running chkdsk to “repair” it. CHKDSK is built to repair a recognized but damaged file system. It generally can’t even run against a volume Windows has already given up on and labeled RAW, and forcing repair attempts on a drive in this state risks writing over exactly the structures a proper evaluation would try to rebuild. If CHKDSK reports it can’t run on a RAW drive, that failure is information, not a dead end. It means the fix needs to happen below the file-system level, not above it. When a drive needs real board or controller-level repair rather than a file-system rebuild, that is a different job at a different price, and it is worth reading what SSD repair can and cannot fix before you commit to either.

The other message worth recognizing: when Windows or macOS offers to format a drive it can’t read, that is the operating system giving up, not a repair option. Declining that offer costs nothing. Accepting it writes directly over the area an evaluation needs.

This is the sharpest version of the corruption-versus-loss distinction on this page. A drive reporting 0 bytes after an unsafe shutdown or power loss has, in most cases, an intact set of NAND flash chips and a controller that simply can’t currently translate between what the operating system asks for and where the data physically lives. Nothing has been erased. The instructions for finding it have been scrambled.

The wrong move here is the same one that shows up everywhere on this page: anything that writes. Initializing the disk in Disk Management, running a manufacturer’s secure-erase or sanitize tool, or updating firmware on a drive already in this state all risk overwriting the very structures that would let the map be rebuilt. None of them are “repair” in the sense of restoring lost information. They start over, which only works if you no longer need what was there.

The right first move is the same one every section here keeps landing on: stop, don’t write to it, and get it evaluated before trying a fix that assumes the drive is simply blank.

Firmware corruption is one of the least visible SSD failures, because from the outside a firmware-corrupted drive and a genuinely dead one can look identical: not detected, or detected with no usable capacity. The difference is entirely internal. A dead controller chip is a hardware problem. Corrupted firmware is a software problem sitting on top of working hardware, and the two need completely different approaches, which is exactly why guessing at a fix from the outside is risky.

A question worth answering directly, because it’s a real and reasonable fear: does updating SSD firmware erase data? It can. A firmware update interrupted by a power loss or a system crash is one of the most common ways firmware gets corrupted in the first place, and attempting a firmware re-flash on a drive that’s already showing corruption symptoms, without the exact matching firmware version and the right tools, can turn a recoverable situation into one that isn’t. Firmware updates are not undone by unplugging the drive partway through. Whatever state they’re in when power drops is the state that sticks.

If a drive stopped responding partway through an update, or you suspect firmware corruption from an unsafe shutdown, the safest next action is to stop powering it on and get it evaluated, rather than retrying the update in the hope that it finishes cleanly the second time.

None of these are bad advice in general. They’re the right fix for a drive with a simple software lock or an ordinary file-system error. The problem is that they are also the first result for a drive with real underlying corruption, where the same command does the opposite of what the guide promises. If you’re not sure which situation you’re in, the evaluation answers that before anything gets written.

The rush fee buys position, not certainty. It speeds up when we start, and it does not guarantee data is recoverable. The evaluation always comes first, you always see what was found before you commit, and No Data Recovered, No Data Recovery Fee means exactly what it says.

You send the drive, we evaluate it at no charge, and you are told what is on it and what it will cost before anything else happens. Nothing is attempted until you say so. That order is deliberate. Evaluate, report, then decide. Nobody should pay to find out the answer is no. If part of what worries you is whether a cloud-synced copy already has you covered, does cloud storage change backup needs is worth reading first.

Recover-SSD was founded in 2011 by Bruce Cullen, who has worked in data recovery since 1999. Drives that arrive untouched come back fuller than drives someone has already tried to fix, and that gap only widens the longer a failing drive stays powered. If your drive matches anything on this page, the useful next step is to stop using it and get the drive diagnosed. The evaluation costs nothing.