Semiconductor ten year lifetime parameter shift tests

[splin]

Whilst trying to understand the long term stability specifications for a voltage reference, I came across this post by a TI engineer which quite surprised me as I've never come across any mention of it before.

http://e2e.ti.com/support/other_analog/etc_analog/f/247/t/343137.aspx

All of our products, including REF5050, undergo a life-time test in a high temperature oven to determine an equivalent shift of various parameters after 10-years (87,600 hours) of constant operation at ambient temperature.  Based on the above test, the expected long-term shift after 10 years of operation is equal or less to the maximum initial accuracy of 0.05% specified in the datasheet (500ppm).

And from the link in that post:
http://e2e.ti.com/blogs_/b/precisionhub/archive/2014/02/20/ic-long-term-stability-the-only-constant-is-change.aspx

    Specifications centered around zero or a mean value (like Vos, Vos Drift, Vref, AOL, CMRR, PSRR, etc.) may shift over 10-year life up to:

                             +/-100% of the max (min) PDS specified value

To be clear, I am not surprised that lifetime testing is done, but by the fact that it is apparently an industry standard test whereby parameters must not shift by more than 100% over 10 years and I was completly unaware of it.

So the question is, is this almost universally known by competent engineers and its just me showing my ignorance?

And in another reponse in the forum by the TI engineer:

It is a semicunductor industry-wide standard to life-test all new products for an equivalent 10-year operation at 25C and there are many industry standards like this that are not specifically mentioned in TI's or our competitors' datasheets.

Can anyone point me to find some more of these industry standards as it seems I might have a missed an important part of my education?

Splin

[MK]

They use the arrenhius equation to tell them what temperature will give them "10 years" of ageing in a week for example.

It does have to be used with caution as some suppliers were using it badly to declare that their voltage references would only age at 1ppm in 10 years or so, clearly wrong on that one.

[David Hess]

They use the arrenhius equation to tell them what temperature will give them "10 years" of ageing in a week for example.

It does have to be used with caution as some suppliers were using it badly to declare that their voltage references would only age at 1ppm in 10 years or so, clearly wrong on that one.

The Arrhenius equation was the first thing I thought of but I assume they temper it with historical data on performance.  I seem to recall reading about the false 1 ppm voltage reference specification in connection with it in ED or EDN or something similar.