19 Mar
2013
19 Mar
'13
6:09 p.m.
On Mar 19, 2013, at 7:13 PM, Liam Proven <lproven at gmail.com> wrote:
From the ad:
They can demonstrate given empirical evidence. Certainly, when
we determine the MTBF of electronic devices, it all comes down to
statistics. If you test a million capacitors over a period of a
year and five of them fail, statistically a given capacitor will
fail every 200,000 years. You can apply various factors to
accelerate the effects of aging that have well-studied time
coefficients; heat is a particularly useful one in electronics
(and, I would assume, media as well).
MTBF analysis is a really interesting field full of exciting
reading! I'd urge people to look into what's done to determine
things like 100-year archival paper testing on paper that's only
been around a few years before they start saying it's impossible.
Of course, as Jim S mentioned, it's totally possible to use
absolutely useless metrics or do the math wrong to come up with
outlandish MTBF numbers. For example, our client somehow came
up with MTBF numbers for their board of over 900 years, which is
almost certainly inaccurate given the quality of the board.
It?s easy. It?s proven. And it?s the new standard
in digital storage.
It's proven, is it? They can actually demonstrate the claim that their
disc will last 1KY? However, this might, of course, merely expose other,
different failure
modes... :?)
There's always that.
- Dave