The drive shows a clear hardware or software fault in its internal controller, which is incorrectly writing nonsensical values to the SMART log (a known issue with some Samsung series). According to the NVMe log, the drive is reporting 91% wear (Percentage Used) after only 300 hours of operation. The drive itself is physically fine, showing no damaged cells (Available Spare is 100%) and no data integrity errors (Media and Data Integrity Errors is 0). However, since the controller is incorrectly reporting 91% wear, there is a real risk that the drive will automatically software-lock itself into Read-Only mode within a few days or weeks once it reaches 100%. According to the output of the nvme smart-log command, the drive reports: Power On Hours: 300 hours (= 1,080,000 seconds) Data Units Written: 7,714,757,057 According to the NVMe specification, 1 Data Unit corresponds to a size of 500,000 bytes (to be precise, 1000 blocks of 512 bytes, so 512,000 bytes for a formatted 512 LBA). Total data volume: 7,714,757,057 × 512,000 = 3,949,955,613,184,000 bytes (i.e. approx. 3.95 PB or 3,950,000 GB). For the drive to physically write 3,950,000 GB of data in 300 hours of operation, it would have to write at a speed of: 3,950,000 GB / 1,080,000 seconds = 3.65 GB/s To achieve these SMART values, the drive would have had to write at a continuous average speed of 3.65 GB/s for the entire 300 hours, from the very first second it was turned on. This is technologically and physically impossible for two reasons: Thermal Throttling: With a sustained write speed of 3.65 GB/s, the drive would overheat within a few minutes. However, the maximum recorded temperature in the SMART data is 40°C, and the "Warning/Critical Temperature Time" shows 0. TLC cache: After the fast SLC cache is exhausted, the write speed for these chips drops to a fraction of that (typically below 1.5 GB/s), which makes a sustained average of 3.65 GB/s over 12.5 days absolutely impossible. Since the physical writing of this volume of data is not possible under the given conditions, and a normal level of writes in the MB/s range was confirmed at the time of inspection, this is a clear reporting fault of the drive itself (Firmware hallucination / Controller bug).