How to Check an SSD: Health, Endurance, and Speed

You can check an SSD by reviewing its SMART data, remaining endurance, and read and write speeds. Use a diagnostic tool, then compare the results with the drive’s specifications.

SSD health cannot be reduced to a single number: errors, wear, and signs of overheating all matter. A speed test can reveal a performance drop, but results depend on the model, connection, and current workload. Let’s look at which metrics to check and how to tell whether they need attention.

• around 500-600 MB/s typical SATA SSD speed
• from 3000 MB/s benchmark for NVMe speed
• about 3.5 GB/s per lane speed limit for PCIe 3.0

How do you check SSD health with SMART?

You can check SSD health with SMART in CrystalDiskInfo: select the drive and review its overall status alongside individual metrics, rather than relying only on the “Good” label. SMART fields can reveal wear and errors that the overall status may not show.

Which metrics to check

• Percentage Used shows what share of the estimated endurance has already been used. The closer the figure is to 100%, the closer the SSD is to its rated endurance limit.
• Available Spare indicates the reserve of spare blocks. Compare it with the threshold set by the drive itself: a value below that threshold needs attention.
• Media Errors counts media errors. If the figure is rising, copy important files and check the SSD again to see whether errors continue to appear.

Evaluate SMART values together: a “Good” status doesn’t remove the need to monitor Percentage Used, Available Spare, and Media Errors. If a metric raises concern, save important data first, then check again in CrystalDiskInfo and see whether the values have changed.

How do you check an SSD in Windows, Linux, and macOS?

You can check SSD health in Windows with PowerShell, and in Linux and macOS with the SMART utility smartctl; CrystalDiskInfo is also suitable for a quick check in Windows. Built-in tools show detected drives and available reliability metrics, while the manufacturer’s software may display the specifications of a particular model more accurately.

• Windows PowerShell: run Get-PhysicalDisk to see the physical drives detected by the system and their status. Use Get-StorageReliabilityCounter to retrieve reliability counters.
• Linux and macOS: view SMART attributes with smartctl. First identify the device name and check whether it supports SMART; then retrieve the drive’s metrics.
• Windows, graphical tools: CrystalDiskInfo is suitable for a quick check. The SSD manufacturer’s utility may show more accurate details for a particular drive.

How to read the results

The fields returned by Get-StorageReliabilityCounter depend on the controller and driver, so some metrics may be missing. Compare the available SMART data with the manufacturer’s description of the relevant metric rather than treating a missing field as a sign of a fault.

How do you estimate remaining endurance from TBW and host writes?

Compare the TBW listed in the specifications for your specific SSD model with the Total Host Writes field, also called “Total Host Writes” or “Host Writes”: this lets you estimate what share of the rated write endurance has already been used. For example, with an endurance rating of 600 TBW and 120 TB of host writes, the estimated share used is 20%.

The estimated remaining endurance in this example is 480 TB before the rated TBW. This is an estimate based on the volume of data written by the host, not a promise that the SSD will keep working until it reaches that point: this metric alone does not determine the drive’s lifespan.

How to interpret Percentage Used

If the utility displays Percentage Used, treat it as an additional indicator and compare it with the estimate based on host writes. The meaning of this field depends on the specific SSD’s methodology, so don’t treat it as a universal measure of remaining endurance.

How do you test SSD speed and interpret the result?

To test SSD speed and interpret the result correctly, run CrystalDiskMark, choose the test size and number of runs, then compare read and write speeds with the drive’s specifications and interface. Close file transfers and resource-intensive apps before testing: concurrent activity changes the result.

How to assess speed

For an initial assessment, compare CrystalDiskMark’s readings with the SSD’s connection type. Here are benchmarks for common interfaces:

• SATA SSD: usually around 500-600 MB/s.
• NVMe: from 3000 MB/s.
• PCIe 5.0: significantly faster than NVMe drives on earlier PCIe generations.

The manufacturer’s utility is useful for checking rated specifications: you can use it to compare the SSD model and firmware with the test result. If the measured speed differs significantly from the specifications, repeat the test without background activity and check that the selected test matches the drive’s interface.

Why might an SSD run slower than expected?

An SSD may run slower than expected because of interface limits, test settings, and the behavior of the drive during sustained writes. PCIe 3.0 is limited to about 3.5 GB/s per lane: a faster drive won’t reach its maximum in a slot using this interface.

What to consider when testing

• PCIe slot. Check that the drive’s interface matches the slot it’s installed in: the SSD cannot exceed the capabilities of the connection.
• CrystalDiskMark. The utility shows the result for the selected test and settings, but doesn’t guarantee the same speed during sustained real-world writes. A short benchmark doesn’t always reflect how an SSD performs when copying large files.
• Repeated write tests. Each run adds writes to the drive. Don’t run them unnecessarily, especially if Percentage Used is already high: it indicates the share of the SSD’s endurance that has been used.
• The wmic diskdrive get model, status command. It displays the drive model and overall status, but doesn’t replace checking SMART fields. To assess health and endurance, check SMART metrics, including “Total Host Writes.”

If benchmark results are lower than expected, first check the connection interface, then repeat the measurement with the same settings. Don’t rely on just one speed test; SMART metrics also help you assess the SSD’s health and the amount of data written.

What should you do if SMART reports a problem?

If SMART reports a problem, first save important files to another physical storage device, then check the warning in a different program. If Media Errors are increasing, don’t run benchmarks or put off copying your data: these errors are a reason not to rely on the SSD as the only place where your data is stored.

How to verify a warning

Don’t base your decision on CrystalDiskInfo alone: compare its status with the readings from smartctl or the SSD manufacturer’s utility. If Available Spare has fallen below the threshold specified for the drive, check this metric in the manufacturer’s software and prepare to replace the drive.

After backing up your data, check the SSD connection and read the SMART metrics again. If errors persist, use the drive only for secondary tasks or diagnostics, and don’t keep the only copy of important data on it.

Frequently asked questions

Key takeaways

• To assess health, check Percentage Used, Available Spare, and Media Errors in SMART.
• Compare Total Host Writes with TBW, but treat the calculation as an estimate, not a prediction of the failure date.
• Use CrystalDiskMark for benchmarking and account for the interface, slot, and background activity.
• If there are media errors, save important files first, then check SMART again with another utility.