Lawrence D’Oliveiro snipped-for-privacy@NZ.invalid wrote: |----------------------------------------------------------------------| |"I found out that a common measure of the projected life of an SSD is | |being given in units of “terabytes written” (TBW). This will typically| |be much larger than the capacity of the drive itself, because it | |includes the sum total of all write operations on the drive, including| |deletion of data. | | | |For example, a 1-terabyte drive could have an expected figure of | |merit, by this measure, of 1000TBW -- that means it should be able to | |endure a total of a petabyte (1000 terabytes) written to it before | |showing signs of failure. | | | |But it seems to me, a 2-terabyte drive made of parts with the same | |quality, meaning each storage cell has the same expected endurance as | |before, should have a proportionately greater figure of merit, namely | |2000TBW. | | | |But the bigger drive will likely not last longer than the smaller one | |-- unless you don’t actually make use of the extra space. | | | |So why not divide the TBW by the actual capacity of the drive? Then | |you end up with a ratio of how much can be written in total, to the | |drive capacity -- call it, say, the “cumulative write ratio”. Both | |those drives would have a cumulative write ratio of 1000. | | | |That number, it seems to me, correlates better to the overall quality | |of the unit than TBW does. E.g. if you see a bigger drive with a | |smaller cumulative write ratio, you can suspect that they are cutting | |corners somewhere to keep the cost down." | |----------------------------------------------------------------------|
I do not yet have a first-hand experience of an SSD failure. Don Y complains in news:sci.electronics.design that an SSD suddenly completely failed, instead of gradual partial degradations of a hard disk. What does Lawrence D’Oliveiro mean by "before showing signs of failure."? This quotation gave me the impression that Lawrence D’Oliveiro does not expect the dramatic all-or-nothing scenario that Don Y reported.
This is a crosspost to news:sci.electronics.design (S.