Check wear, host writes and temperature before an SSD fails - in CrystalDiskInfo and with the counters Windows already exposes.
**CrystalDiskInfo** shows health, host writes and temperature in one window; **Get-StorageReliabilityCounter** gives the same numbers in PowerShell. Above **80 °C** most NVMe drives throttle.
Start CrystalDiskInfo and select the drive. Health Status, Temperature, Total Host Writes, Power On Hours and Power On Count are the numbers that matter.
Compare Total Host Writes with the TBW rating from the vendor's spec sheet. 40 TB written on a 600 TBW drive is under 7% of its rated life.
The reliability counters give temperature, wear and error counts without third-party tools. Blank fields are normal on some NVMe drivers.
Get-PhysicalDisk | Select-Object FriendlyName, HealthStatus, MediaType
Get-PhysicalDisk | Get-StorageReliabilityCounter | Format-List Temperature, Wear, ReadErrorsTotal, WriteErrorsTotalCopy a large folder or run the vendor's self test and watch the temperature. Most consumer NVMe drives start throttling near 80 °C.
In CrystalDiskInfo, Reallocated Sectors, Pending Sectors and Uncorrectable Errors should all be zero. Any value that rises between two checks means back up today.
Note the date, host writes and idle temperature. The next check compares numbers instead of memory, and a jump in writes tells you which app is doing the damage.
The drive sits behind RAID or Intel RST. Switch the controller to AHCI or read it with the vendor tool.
The driver does not expose them. Use CrystalDiskInfo or the vendor utility instead.
No airflow over the drive. Check the heatsink contact and lower the package power limits; a thermal pad is often enough.
For short bursts no, sustained yes. Throttling and faster wear both start in that range.
A short self test once a month is a good habit; the SMART counters matter more.
Nothing to undo - both tools are read-only. Delete the portable CrystalDiskInfo folder when you are done.
Free · no account needed · tracks your progress