[svlug] archive media of choice for 10~100GB

[John Conover]

Ooops! I stand corrected. Change 8002 to Mil. Std. 883.

Sorry about that.

        John

John Conover writes:
> 
> See:
> 
>     http://www.dscc.dla.mil/Programs/MilSpec/SmdOverview.html
> 
> about Mil. Std. 38510 for starters. Google will find others.
> 
> Almost all space electronics, which is designed for a total unattended
> system operational life of 25 years, (meaning the theoretical MTBF is
> 5-10 times that,) references the 38510 document for the QA methodology
> of the HW/chips.
> 
> Note that the failure model of 38510 for electronic stuff assumes
> ergodicity, (e.g., the century failure rate can be determined by
> measuring the failure rate of a hundred systems for a year.) It isn't
> so-as pointed out by Karen-but its a "reasonable" stochastic modeling
> assumption, based on a 2 year QA process, and a way to do things that
> has a "reasonable" documented history in the industry. I mean, Voyager
> is still beeping at us from the edge of the effects of the sun in deep
> space, and has never had a Ctrl-Alt-Del.
> 
> As rough cost numbers for 38510/8002-A, multiply the consumer cost at
> Fry's by somewhere between a hundred and a thousand X, (more for the
> 8002 "man rated S" space spec.)
> 
>         John
> 
> Chris Palmer writes:
> > 
> > Karen Shaeffer writes:
> > 
> > [entropy effect on magnetic and solid-state stuff]
> > 
> > Is there any chance you can cite a reference for this, or is there some=20
> > basic textbook that explains it all? Thanks.
> > 
-- 

John Conover, conover at rahul.net, http://www.rahul.net/conover/