[time-nuts] Rubidium standard / MTBF

Lux, Jim (337C) james.p.lux at jpl.nasa.gov
Thu Nov 19 15:17:01 UTC 2009




On 11/19/09 4:56 AM, "Bob Camp" <lists at cq.nu> wrote:

> Hi
> 
> The one that I find the most shocking is the very major internet hardware
> company that considers 5 years of continuos use to be the goal. The logic - we
> want them to swap out the gear regularly....

That's actually reasonable.. Moore's law is always advancing, performance is
improving, and regular swaps avoid needing to support old versions for long
times.  A pretty simple financial analysis shows what the optimum plan would
be, and I'm sure that's what the MIHC does.

We face the opposite issue at JPL.. Supporting 20 year old hardware in space
(which is pretty benign) or more to the point some ground testbed that
replicates it.  The flight hardware or prototype probably isn't the problem,
it's the Apple II computer hooked up to it in the testbed that gives you
diagnostic information, or is used to generate the software builds using a
program that runs on the Microsoft BASIC card installed in that Apple.  (or
Rocky Mountain Basic on that HP calculator, or...)




More information about the time-nuts mailing list