Drive Health Calculator
Enter your S.M.A.R.T. readings and get a plain answer: healthy, worth watching, or plan a replacement. The calculator runs entirely in your browser. Nothing is uploaded and no account is needed.
You need the raw values for a handful of attributes. Any tool that reads S.M.A.R.T. will show these — the drive's own firmware reports them, and the tool just passes the number through. If you have a second set of readings from a few weeks ago, add them; trends improve the assessment more than any single snapshot can.
Open the attribute databaseRun the assessment
Every field is optional except the drive type. Enter what you have — the assessment explains which input drove the result, so a rough entry still gives you a readable answer.
Runs in your browser
The calculator does not read your drive directly and does not connect to your system. Values you type are not transmitted or saved, and there is nothing to sign up for.
The output names the signal that decided the result, so you can check the reasoning instead of trusting a label. Three outcomes are possible.
No tracked counter is out of range. That does not mean the drive will last forever — it means the numbers it reports today look normal for its age and type. Keep a copy of the readings somewhere; the next comparison is where the real information lives.
At least one signal is worth rechecking, usually every two to four weeks. A pending sector that clears on its own reads very differently from one that sits there for a month, and one extra reallocated sector in a fresh drive is not the same story as ten new ones in a week.
Treat a watch result as a reason to have a current backup, not a reason to panic.
At least one warning is active and either unresolved or growing. The output names which input caused it, so your first move is checking that attribute's entry in the database rather than shopping for a drive. Backing up comes before anything else.
This is how hard drive failure indicators actually behave. Read the numbers the way a technician does — as a direction of travel, not a photograph.
The calculator weighs unresolved warnings more heavily than resolved ones. Pending sectors, uncorrectable errors, and a growing reallocated count push the result toward watch or replace. High power-on hours alone rarely do.
That ordering follows how drives tend to behave in the field. A six-year-old drive with a clean surface-error record is a different animal from a two-year-old drive that began logging pending sectors last month — even though the second one has far fewer hours on it.
The output explains which input drove the result, so the reasoning is visible and you can disagree with it.
| Reading you supply | Weight in the result | What pushes it higher |
|---|---|---|
| Current pending sectors | Heaviest unresolved signal | Any count that stays after a recheck, or rises between two readings |
| Uncorrectable errors | Heavy when it climbs | Growth between readings, especially alongside a clean surface record |
| Reallocated sectors | Moderate, growth-weighted | A rising raw value; a stable value near zero contributes little |
| Temperature | Context, plus a hard high-side ceiling | Sustained readings well above the operating range your vendor documents |
| Power-on hours | Light — age context only | Very little on its own; a continuously running NAS drive is built for it |
| SSD life left | Heavy for solid state drives | A low reported percentage, read with the vendor's own endurance guidance |
Weighting reflects common field behavior, not a vendor warranty rule. For the exact documented operating range of your model, check the manufacturer's own datasheet.
Hard drives are judged mostly on surface errors and mechanical wear. Solid state drives are judged mostly on endurance metrics like wear level and life left, plus controller-reported errors. The calculator adjusts its weighting when you tell it which drive type you have. Skipping that step gives a rougher estimate.
Spinning drives
Solid state drives
A single snapshot can mislead. One reallocated sector logged during manufacturing looks identical to the first sign of surface damage until you see whether it grows.
If you can provide an earlier value, the calculator compares the two and flags growth specifically. That difference between a stable number and a moving one is the single most useful thing a health assessment can tell you, and it is the reason the two optional "earlier" fields sit in the form at all.
No earlier reading handy? Note today's values, then recheck in two to four weeks. The second snapshot costs you nothing but time.
Solid state wear is reported differently by each vendor. Crucial documents remaining-life indicators; Samsung surfaces endurance through Magician; Kingston and others use their own ID ranges. If your tool shows a life-left percentage, enter it as-is and note the vendor so the weighting is accurate.
A number your tool calls "wear level" may be a count, a percentage, or an inverted scale depending on the vendor. Reading the manufacturer note for your brand before you compare two drives saves a lot of confusion.
Read the manufacturer notesIt does not read your drive directly, does not connect to your system, and does not replace a backup. It interprets numbers you supply.
It also cannot see mechanical damage that has not yet been logged, and it cannot predict the future. A healthy result is a statement about the counters you entered today. If a result says replace, the next step is copying data off the drive, not rerunning the calculator.
Read the dying-drive guide| Attribute | Typical healthy range | Where it starts to matter |
|---|---|---|
| Reallocated sectors count | A small non-zero value that never changes | Any steady increase between readings |
| Current pending sectors | Zero, or a count that clears on the next read | A count that persists across a recheck |
| Uncorrectable errors | Zero | Any value that grows |
| Temperature | Comfortably inside the vendor's stated operating range | Sustained readings near the top of that range |
| Power-on hours | No universal ceiling; depends on drive class | Only when paired with surface or endurance warnings |
| SSD life left | High remaining percentage on the vendor's own scale | A low figure on that scale, read with the vendor note |
Ranges above describe common field behavior. Vendors publish their own limits and those override general guidance — start from the database entry for the attribute, then check your model's datasheet.
Back up first. Then confirm which attribute triggered the flag by rereading the database entry for that ID.
If the only signal is a rising uncorrectable error count with a clean surface-error record, check the SATA or NVMe cable and reseat the drive before replacing it. Cables fail more often than people expect, and a reseat costs nothing.
Read the dying-drive guideShort answers to the four things people ask most often before they trust a number from a browser tool.
No. It runs entirely in your browser. Values you type are not transmitted or saved. The optional email field is the only thing that leaves the page, and it is used to reply to you, nothing else.
Yes. Select the drive type so the weighting matches HDD surface-error logic or SSD endurance logic. An NVMe drive reports a different set of endurance counters from a SATA model, which is why the selector separates the two rather than lumping every solid state drive together.
You still get an assessment, but trend-based warnings need a second reading. Recheck in two to four weeks if anything looks close to a boundary. Write the raw numbers down somewhere you will find them again.
Hours alone rarely predict failure. A NAS drive in a Toshiba N300 is built to run continuously. Surface errors and unresolved warnings matter more. Hours tell you how much time the drive has spent working, not how close it is to the end of its useful life.
The calculator gives you an answer about today. Save the raw values somewhere — a text file, a note, the back of an envelope — and run it again after a few weeks of use. That second run is where a drive health check starts telling you something a single reading cannot.
Questions about a specific result? Call +1-408-555-0192, Monday through Friday, 9:00 AM to 6:00 PM Pacific Time, or write to info@smartdrivehealt.live. Office: 452 Silicon Drive, Suite 102, San Jose, CA 95110, USA.