[wildzzz]

This is likely just something that corporate users would care about. Companies often lease PCs from IT service providers rather than own and maintain their own hardware. The owner of the hardware now has a metric they can point to for how "usable" a machine is after the initial lease. As a customer, I may not want to rent laptops that have been through who knows what sort of wear and tear no matter how cheap but if the owner can now show me actual data saying how used the laptop is, I may feel more comfortable paying less for used. It's like the odometer on a car, I'd never buy a used car that didn't have an odometer (even if such a thing existed). But with an odometer, I can get a general idea of how much use a car has had despite the age. Only a year old with 30k miles? Hell no. Three years old with only 10k? That car might as well be new.

I'm assuming since it writes to a vendor-reserved sector, replacing it would make the whole thing moot. The rental company wants to retain that data because it makes a used PC more valuable. Since the corporate renter doesn't own the PC, they would only be allowed to wipe the SSD (excluding this section), not remove and destroy it.

[buccal]

"Odometer" for HDDs and SSDs are already provided in SMART data that is more or less standatized and accessible using many tools. The data is not resettable by mortals similarly to car odometers.

[aspenmayer]

Everyone should test their hard drives within the return window after purchase for these kinds of errors as it is a warning and failure condition, but there is no reason to buy a caution or warning status drive. The metrics are standardized somewhat but the interpretations of the values are vendor and model specific. I’ve used CrystalDiskInfo in the past but don’t know what the state of the art in this is at the moment and this isn’t a recommendation per se. I just have backed up a lot of hard drives, and when I have block or file level access issues, there are usually SMART errors, but not always on the vendor provided tools. Sometimes the third party tools are wrong, but I had a pretty large sample size of computers that had unknown problems which is not likely representative of most computers or computer users, but the tools were helpful as far as knowing more about the hardware to inform the user, so I can see where HP is coming from. I suspect some kind of gaming of the metrics but can’t prove this. It’s not a horrible idea, but I don’t love more failure modes. I’ve seen BitLocker issues from updating firmware without turning it off first, which is cautioned against as a possibility, but if this helps with those kinds of issues too, it will be sold as a feature for the corporate market to help with fleet management.

[wildzzz]

Sure but that doesn't apply to the whole machine. What HP is proposing is SMART data for much more than just the storage device. The only thing that would make this better is if the data can be stored in NV flash on the motherboard rather than a vendor-reserved sector on the SSD. This way failed drives can be replaced while preserving the data.