Windows recently warned me about the SSD in my laptop, and a health check now reports only 3% health with a status of "Bad." The 512 GB NVMe drive is a Phison ESO512GHLCA1-E9C, running at 53°C. It has been powered on for about 6,652 hours and reports roughly 44 TB read but 1,093,713 GB written—over 1 petabyte—in less than two years. The SMART data also shows a critical warning, 61% percentage used, 182 unsafe shutdowns, and 35 error-log entries, although there are currently no media or data integrity errors. I have never knowingly used hibernation. Is the drive genuinely close to failure, why might the write total be so high, and why did Windows wait until 3% to warn me?
4 Answers
Phison is the controller manufacturer, so the full model number is useful when checking the drive’s rated endurance and firmware. The SMART values suggest the drive has experienced substantial wear, even though it has not recorded media errors yet. Also leave around 10% of the replacement drive empty when possible; that can help performance and endurance, but it will not restore worn cells on this SSD.
Treat the drive as unreliable now. Back up or clone anything important immediately, then replace the SSD rather than waiting for it to fail. A health percentage is an estimate based on the drive’s endurance counters, and once the monitoring software reports a critical warning, it should not be trusted for important data. Keep the replacement installation fresh instead of continuing to use this drive as the system disk.
Do not rely on recovery software as your backup plan. Copy irreplaceable files first, preferably to two separate destinations, and avoid unnecessary restarts or heavy disk activity until that is done. Afterward, investigate the unusually high host-write total and check whether the laptop manufacturer has a firmware or SSD replacement program.
The write figure is the biggest clue. More than 1 PB written to a 512 GB drive means something has been repeatedly writing large amounts of data—possibly crash logs, a sync or backup program, an update loop, virtual machines, torrents, browser caches, or another runaway process. Check Windows Resource Monitor or a disk-usage monitor after replacing the drive so you can identify which process is generating the writes. Hibernation is one possibility, but it is not the only explanation, and it would not repair the worn-out drive.
The temperature is not especially alarming by itself, and a lack of obvious slowdown does not prove the drive is healthy. SSD wear counters and failures can appear without much performance warning.
That makes sense. I was mainly surprised that the first warning appeared at 3% instead of giving me a warning earlier.