Storage · Guide

Is my SSD dying?

A drive health tool shows you forty numbers. Four of them matter, and none of them is a promise.

By Marcin Firmuga·Published 2026-07-29·8 min read

You install a drive health tool because something felt off, and it hands you a wall of forty SMART attributes with names like "Seek Error Rate" and raw values in the millions. Somewhere in there is the answer to the only question you have: do I need to buy a drive this week or not?

Good news, most of that table is noise for your purposes. A handful of values genuinely predict failure, a few more are famously alarming and completely harmless, and the whole report comes with a catch worth knowing before you trust it. Here is how to read your drive without a computer science degree, and how PC Workman boils it down to a verdict you can act on.

In this guide
  1. The values that actually matter
  2. SSD wear, and why it worries people more than it should
  3. The symptoms that beat any number
  4. The catch nobody mentions
  5. How PC Workman reports it

The values that actually matter

Different drive types report health differently, so start by knowing which kind you have.

On an SSD or NVMe drive, two values carry most of the signal. Percentage Used (or Media Wearout Indicator) estimates how much of the flash endurance you have consumed, and it counts up from 0. Available Spare shows the reserve blocks the drive keeps to replace worn ones, and it counts down from 100 percent. Spare dropping toward its threshold is the serious one. Also worth a glance: any Critical Warning flag, and Media or Data Integrity Errors, which should stay at zero.

On a mechanical hard drive, the classics still hold. Reallocated Sectors Count counts sectors the drive gave up on and replaced from spare. Current Pending Sector Count counts sectors it cannot read reliably and has not yet reassigned. Uncorrectable Sector Count counts the ones it failed on outright. All three should be zero. Any of them rising, especially over weeks, is the clearest software warning a hard drive gives you.

Ignore the ones that scare everyone. On many Seagate drives, Raw Read Error Rate and Seek Error Rate show enormous raw numbers by design, because the manufacturer encodes several counters into one field. A raw value of 180 million there is not 180 million errors. Judge these by the normalised value against its threshold, or better, ignore them and watch the four above.

SSD wear, and why it worries people more than it should

Every SSD ships with an endurance rating, usually expressed as TBW (terabytes written). A typical consumer 1 TB drive is rated somewhere in the hundreds of TBW. That number sounds finite and frightening until you compare it with how much you actually write.

A normal desktop user writes a few terabytes a year. Even a heavy one, with games installed and removed constantly and large video files moving around, rarely gets close to the rating inside the drive's practical lifetime. Percentage Used climbing by a couple of points a year is completely ordinary. If yours is at 4 percent after two years, you are on pace to replace that machine long before the flash gives out.

Wear is worth watching, not worrying about, and it is one of those numbers where the trend across months tells you far more than today's reading, exactly like reading a voltage rail over time rather than panicking at one sample.

The symptoms that beat any number

Sometimes the drive tells you before SMART does. Treat these as urgent regardless of what a health tool says:

The catch nobody mentions

A clean SMART report is not a promise. Large-scale studies of drive failures have consistently found that a meaningful share of drives fail with no prior SMART warning at all, because controller and electronic faults can kill a drive instantly with nothing to gradually count. SMART is an early warning system for the failures that happen slowly. It cannot see the ones that happen all at once. Which means the health check is worth running, and your backup is the thing that actually saves you.

So the practical rule: if the numbers turn, act early and calmly. Copy your data off first, then decide about replacing. Do not run long stress tests on a drive you suspect, because the test is itself the load that finishes it off.

How PC Workman reports it

The gap between a forty-row SMART table and "should I be worried" is where most people give up. PC Workman reads the values that matter for your specific drive type and states the verdict in plain language first, with the readings underneath:

Drive HealthSamsung SSD, NVMe
Percentage used4%
Available spare100%
Critical warningsnone
Media errors0
Power-on hours7,180
Wear trendabout 2% per year at your usage
healthyNo action needed. Wear is normal for the hours on this drive.

And when something does turn, the honest version says what it can and cannot know: which value moved, over what period, and that a passing check is never a substitute for a backup. That is the same boundary we hold everywhere in the app, including the five-minute health check: report the evidence, name the trend, never fake certainty.

Know your drive before it surprises you. PC Workman reads the health values that actually matter for your drive type, tracks the wear trend over time, and gives you a plain verdict instead of a wall of raw SMART numbers. Free, open source, 100% offline. Download for Windows.

Next question: the drive is fine, so why does the machine take a minute to boot? Browse the rest of the guides →

MF

Marcin Firmuga

Developer of PC Workman · HCK_Labs

18 months of building a system monitor in public. Every guide is based on what the sensors actually report.