SSD Failing? 7 Warning Signs Every Edinburgh PC Owner Should Know

How to spot a dying solid-state drive before it takes your photos, documents and work with it.

29 April 2026 10 min read Hardware Alex M.
SSD Failing? 7 Warning Signs Every Edinburgh PC Owner Should Know

Solid-state drives have been a huge upgrade over old mechanical hard drives — they're faster, quieter, and far more shock-resistant. But SSDs are not immortal. Every flash cell has a limited number of write cycles, and when an SSD does fail it tends to fail suddenly, with very little of the warning noise an old spinning disk would give you.

If your computer is starting to behave oddly, you owe it to yourself to find out whether the drive is the cause. Here are the seven most common signs we see across Edinburgh — from Leith and Stockbridge to Corstorphine, Musselburgh and Penicuik — that an SSD is on its way out.

1. Files Suddenly Become Read-Only

One of the earliest warning signs of a failing SSD is the operating system flipping the drive into read-only mode. You can still open documents and applications, but you can't save changes, install updates, or move files onto the drive. Windows might show errors like "the file system is read-only" or "operation could not be completed".

This is the SSD's controller protecting whatever data is left. If you see this, stop using the drive for new work and copy your important files off straight away.

2. Frequent Freezes, Crashes and Bad Blocks

If your PC freezes for several seconds when you try to open a file, or you start seeing "bad block" or "CRC" errors in Event Viewer, the drive is struggling to read or write to certain cells. On Windows 11 you might also notice the dreaded Blue Screen of Death appearing more often, especially during boot or while saving large files.

3. The Drive Disappears From the BIOS

An SSD that vanishes from the BIOS, comes back after a reboot, then vanishes again is a classic sign of a dying controller chip. NVMe drives in particular can fail this way — Windows boots normally one day, then on the next reboot the system reports "no bootable device". If this is happening intermittently, treat the drive as terminal and back it up immediately.

4. Boot Times Have Crept Up

SSDs are supposed to boot Windows in seconds. If your laptop now takes a minute or more to reach the desktop, or hangs on the spinning dots, the drive may be reaching the end of its life. A few other things can cause a slow boot — too many startup apps, a failing battery, dodgy Windows updates — but the steps in our guide to speeding up a slow Windows PC will help you rule those out before pointing the finger at the SSD.

5. SMART Health Status Is Warning You

Every modern SSD reports its own health through a system called SMART. Tools like CrystalDiskInfo (free) read these values and give you a quick "Good", "Caution" or "Bad" verdict. The numbers worth checking are:

  • Percentage Used / Wear Level — how much of the drive's write endurance you've consumed.
  • Reallocated Sectors / Bad Blocks — cells the controller has marked as unusable.
  • Available Spare — the pool of spare cells used to replace bad ones. When this gets low, failure is close.

A "Caution" rating doesn't always mean immediate death, but it's the universe handing you a free warning. Take it.

6. Files Are Becoming Corrupted

Photos that won't open, Word documents that crash on save, video files that suddenly stutter halfway through — when individual files start corrupting for no obvious reason, the SSD's flash cells are losing their ability to hold data reliably. This is a particularly nasty failure mode because backups taken after the corruption starts may already contain damaged copies. If you spot this early, our data recovery service can often pull clean versions from the unaffected sectors.

7. Your PC Refuses to Boot at All

The final stage. The system powers on, the fans spin, but you get a black screen, a flashing cursor, or a "no operating system found" message. If you've ruled out the obvious culprits in our PC won't boot guide and the drive isn't appearing in the BIOS, the SSD has likely failed completely.

The good news: even at this stage, data recovery is sometimes possible if the controller chip can be coaxed back into life or the NAND flash read directly. Don't keep powering the system on and off — that often makes things worse.

Why an SSD Fails Differently From a Hard Drive

A mechanical drive usually gives you warning. It clicks, it whirrs, it slows down for weeks before it finally stops, and by then most people have noticed something is wrong. An SSD has no moving parts, so it has no way of making a noise — and that removes the one early warning most people rely on.

It also fails in two quite separate ways, and they are worth telling apart. The first is wear: every flash cell can only be written a finite number of times, and once the drive has used up its spare cells it starts refusing writes. That is the slow, predictable failure the endurance ratings describe, and honestly, most consumer drives never reach it. The second is the controller — the small processor that decides where every block of data physically lives. When that fails, or its power circuitry does, the drive does not degrade gracefully; it disappears mid-sentence. This is why a two-year-old SSD with 4% wear can die without a single warning sign, and why “it is nearly new, so it is fine” is not a safe assumption. Our comparison of NVMe and SATA SSDs covers how the two interfaces differ when things go wrong.

Reading a SMART Report Properly

The summary word at the top of CrystalDiskInfo — Good, Caution, Bad — is a starting point, not the answer. It is worth knowing which individual values actually matter, because a drive can read “Good” while one number quietly falls off a cliff.

On a SATA SSD, the ones to look at are:

  • Reallocated Sectors Count (05) — cells retired because they could no longer be trusted. A handful is normal over a drive's life; a number that climbs week on week is not.
  • Wear Levelling Count (AD) and SSD Life Left / Percentage Lifetime Remaining (E7) — how much write endurance has been consumed. These count down from 100, which catches people out.
  • Total Host Writes (F1) — the total data written over the drive's life. Compare it against the manufacturer's rated endurance for your capacity to see whether wear is genuinely a factor.
  • Uncorrectable Error Count (C6) — errors the drive could not fix on its own. Anything above zero deserves a backup that same day.

On an NVMe drive the names are different and, in our experience, more useful:

  • Critical Warning — should be 0x00. Anything else means the drive is already telling the operating system it is in trouble.
  • Percentage Used — endurance consumed, counting up. It can legitimately pass 100% and keep working, but it is no longer a drive to keep anything irreplaceable on.
  • Available Spare against Available Spare Threshold — the single most useful pair of numbers on a modern SSD. When the spare pool falls towards the threshold, the drive is close to running out of cells to substitute in.
  • Media and Data Integrity Errors — data the drive could not return correctly. This should be zero, permanently.
  • Unsafe Shutdowns — a high count on a desktop points at power problems rather than the drive, and power-loss during a write is one of the commonest ways a healthy SSD gets corrupted.

One thing to avoid: wmic diskdrive get status reports “OK” on almost everything, including drives that are visibly dying, so it is not a health check. PowerShell's Get-PhysicalDisk | Select FriendlyName, HealthStatus, OperationalStatus is better but still coarse. The manufacturer's own tool — Samsung Magician, Crucial Storage Executive, WD Dashboard, Kingston SSD Manager — reads the same values with the right labels for your model, and is worth installing before you need it.

Firmware: the Failure That Isn't Really a Failure

A recurring category we see in the workshop is the drive that has not failed at all — it has hit a firmware bug. There have been several well-documented cases across different manufacturers where drives stop responding at a specific power-on-hours figure, or after a certain volume of writes, regardless of how healthy they are. The pattern is distinctive: a batch of identical drives, all fine, all suddenly not.

The practical lesson is to check for a firmware update while the drive is still healthy, using the manufacturer's tool. Once a drive has already stopped responding, the update usually cannot be applied, and what was a five-minute job becomes a recovery one. If you are about to clone an old drive onto a new SSD, update the new drive's firmware first, before anything is copied onto it.

What Event Viewer Is Telling You

Windows keeps a record of every time a drive misbehaves, and it is more honest than the machine feels. Open Event Viewer, go to Windows Logs › System, and filter by the sources disk, Ntfs and storahci or stornvme. The events worth knowing:

  • disk, Event ID 7 — a bad block. The drive could not read a sector.
  • disk, Event ID 11 — a controller error. Often the cable or connection on a SATA drive, and often the drive itself on an NVMe one.
  • disk, Event ID 153 — an I/O operation was retried. A few of these under very heavy load are not alarming; a steady drip during ordinary use is a drive struggling.
  • Ntfs, Event ID 55 or 98 — file system corruption detected. Worth backing up before doing anything else.

The useful part is the timeline. Errors that started three weeks ago and are becoming more frequent tell you far more than a single entry, and they give a technician a much better idea of whether you are looking at a drive on the edge or one that simply had a bad afternoon.

The First Ten Minutes Matter

If you have read this far because something on this list matches your machine, the order you do things in genuinely changes the outcome:

  1. Stop writing to the drive. Close what you can and resist the urge to keep working while you think about it. Every write to a struggling SSD is another chance for the controller to fail mid-operation.
  2. Do not run repair tools. chkdsk /f, defragmenting, “optimise”, or a reformat-and-try-again all rewrite large areas of the drive. On a marginal SSD they are the single most common way people turn a recoverable situation into an unrecoverable one.
  3. Copy the irreplaceable things first. Photos, documents, anything not backed up elsewhere — in that order, not alphabetically. If the drive is in read-only mode, that is the controller granting you one last window; use it.
  4. If the data really matters, image the whole drive before attempting anything else, and work on the copy. Our data recovery guide explains why imaging first is the professional habit.
  5. Do not keep power-cycling a drive that vanishes and reappears. Each cycle is another opportunity for a controller that is already failing to stop coming back at all. We wrote up exactly this pattern on a portable SSD that kept dropping off.
  6. Get a SMART report saved before anything else is attempted. It tells a technician how much life is left and whether recovery is realistic, and it is much harder to obtain once the drive has stopped responding entirely.
  7. Then plan the replacement. A modern NVMe SSD is a genuine hardware upgrade rather than a like-for-like swap — but do it after the data is safe, not before. Replacement drives are listed in our shop under SSDs and hard drives.

How We Can Help

At PC Repair Services Edinburgh we deal with failing SSDs every week — from MacBook Pros in Bruntsfield to gaming rigs in Dalkeith. We can run a full SMART diagnostic, recover what's still readable, and fit a new drive with a fresh, properly-licensed copy of Windows or macOS so you're back up and running fast.

We offer in-shop work, our home and office callout service across Edinburgh and the Lothians, and remote help where it makes sense. Book a diagnostic or get in touch and we'll let you know the next step.

Last updated: 23 September 2026

Frequently Asked Questions

What Edinburgh customers ask when an SSD starts behaving strangely.

Most consumer SSDs are rated for far more writing than a normal person will ever do — typically well over a hundred terabytes written for a mid-size drive, which is years of ordinary use. In practice, wear is rarely what kills them. Controller failures, power-loss events and firmware faults account for most of the dead drives that reach our bench, and none of those follow a schedule.

Not necessarily, but it means one attribute has crossed a threshold and you should find out which. Sometimes it is a single reallocated sector from years ago that has not changed since; sometimes it is the spare pool dropping. Open the detail view, note which value is flagged, and check it again in a week. A number that is moving is the one to worry about.

Usually not, and that is deliberate — the controller has locked writes to protect what is still readable. Treat it as a fixed amount of time to get your data off rather than a fault to fix. Copy the irreplaceable files first, do not reformat it “to reset it”, and plan on replacing the drive rather than returning it to service.

Sometimes, and it depends entirely on which part failed. If the controller has gone but the flash chips are intact, the data physically still exists and specialist work can often reach it. If the cells themselves have worn out or the drive has done an internal secure-erase after a firmware fault, there may be nothing left to read. It is also much harder than on a mechanical drive, which is why the backup advice matters more with SSDs, not less.

If the data matters, no. Shut it down and leave it. Every hour of use is more writes, more heat, and another chance for a marginal controller to stop responding for good. If you need the machine for work in the meantime, tell us when you book — getting the data off first and worrying about the machine second is nearly always the right order.

Open Task Manager, go to the Performance tab and click the disk — the model name is there, and NVMe drives usually say so. For more detail, msinfo32 under Components › Storage › Disks lists every drive by model. Our guide to checking your PC specs on Windows 11 covers both routes, and the answer matters because the two report their health under different names.

Not on age alone. What is worth doing is checking the numbers once or twice a year, keeping the firmware current, and making sure there is a backup that does not live inside the same machine. A drive that is reporting a full spare pool and no errors is doing its job, however old it is; a drive with no backup behind it is the actual risk.

Worried Your SSD Is on Its Way Out?

Don't wait until the drive disappears completely. A diagnostic now is much cheaper than full data recovery later.

Book a Diagnostic Contact Us