Does your Windows / favorite video game just get corrupted all of a sudden ?
In: Technology
Not unless the SSD controller screws up big time.
SSDs keep track of how many times they write to which block. The manufacturer decides how many time a block is supposed to be written to before it’s considered degraded. When a block reaches its maximum number of writes, the controller sorta *retires* that block and uses a healthier spare block in its place.
Some SSDs have Error Correction (ECC) and spare blocks allocated in case of internal issues. The SSD will generally warn the OS before failure.
Unless you buy cheap hardware…
Depending on the firmware settings it can get into anwrite protected mode so you dont loose your data. However you should get SMART warnings from the BIOS when it gets critical ar POST.
You can use Crystal Disk Info to monitor the healt status on most common SMART attributes.
I think most of the time before you reach the end of life on a SSD it will fail miserably before that.
Also most SSD have overprovosioning, a pool of reserved flash used when the controller starts detecting bad cells it will mark them as unusable if that’s the term and enable new cells to be used.
That’s why you see SSD advertised as 480GB and 500GB.
Look in the datasheet for TBW to get an idea of how much data can be written to the SSD.
For an average home user IIRC it is about 3 to 5 years.
Like some other commenters mentioned, modern SSDs keep a bunch of spare cells. When a cell looks like it is having problems, or has been written to enough times, the firmware in the drive will swap in one of the spare cells.
Modern drives also do something called “wear leveling” where they’ll proactively swap data between cells that have been written more times and ones that have been written fewer times.
But I can anticipate your next question: what happens when you run out of spare cells? And everything has been written to a lot?
Well, there’s a thing called SMART monitoring, where the drive can tell your operating system that it’s having problems. Your OS may surface a lack of spare cells as some kind of warning to you, or there may be a utility to manually check on the health of the drive. At that point you’d want to replace the disk.
Okay, but — what if you ignore that and keep using it?
Flash memory works by putting a high voltage through a semiconductor memory cell and flipping it between two states (or sometimes three or more, in some newer designs). At some point the cells won’t reliably be able to stay in the correct state, which will manifest as reading back a 0 when you originally wrote a 1 (or vice versa). Depending on the design of the drive there may be some amount of error correction built in, where they store some redundant checksum data that enables the disk controller to still return the correct data even if a small number of bits on the storage medium are flipped.
But eventually there will be so many flipped bits that either the checksum fails, or the data is wrong but the checksum happens to be right. In the former case, your operating system will report some kind of error when attempting to read the file. In the latter case, you’ll read back “corrupted” data that is different from what was originally written. If the drive only contains data files, you just won’t be able to read them, or they will contain increasing amounts of garbage data. At some point some of the blocks containing filesystem metadata will become unusable or corrupted, which will likely result in the whole drive being inaccessible. If system files become corrupted the operating system will likely stop working properly, possibly “bricking” the computer.
If you are lucky, yes. If you are unlucky, your SSD will suddenly stop working *entirely* with zero warning.
Latest Answers