[kwlug-disc] SSD life

[Remi Gauvin]

On 2026-06-23 2:38 a.m., Ron Singh wrote:
> It's the literal age of the SSD I am really worried about -- 9+ years
> since mfg, 2.5yr of power-on hours. 
> Cells holding an electrical charge reliably after 10 years say, how
> much faith should I have in that?


All of this is wisdom I have gathered from the internet and my own
personal experience. I am not an engineer, I just play one on the internet.

New Cells can reliably hold a charge for approx 10 years, *at rest*.
This would be a grave concern if the ssd was left sitting on a shelf. An
in-use device, however, will have re-written those cells several times
over with regular wear leveling.

(As the drive goes through more re-write cycles, the expected length of
time it can hold a charge will decrease.  A consumer drive should still
be able retain data for 1 year when it reaches the rated number of
re-write cycles. )

Of course, electronics fail on an age bathtub curb, so I can understand
why you would be wary of a 9 yr old device.  To that, all I can say, I
have experienced *far* more ssd failures on drives <1 yr old than I have
on drives over 10 years old, so I would not swap out a perfectly working
drive yet. I don't have actually good statistics to back that up with,
(and really, since we would have to wait over 10 years for each new
model or generation of of NVRAM, I don't think we ever will.)