[ExI] Why space tech isn't cutting edge

spike spike66 at att.net
Mon Nov 19 17:22:19 UTC 2012

>... On Behalf Of BillK
Subject: Re: [ExI] Why space tech isn't cutting edge

On Mon, Nov 19, 2012 at 4:21 PM, Dan wrote:
>>... Why is this news? I thought it was widely known that electronics on
space missions is always several years behind what's being done elsewhere.

>...It is not just several years. It's more like 15 years behind.

Depends on how you count it.  If we get hit with a nuclear EMP, those space
chips will survive, while all our modern stuff and the ability to make them
all will be up in smoke.  Then the i486 will be 15 years ahead.

>...Your ordinary pc chips can't just be sent up into space. The radiation
out there would cause them to deteriorate / malfunction quite quickly, as
the article explains. And no, wrapping them in tin foil doesn't work...

We never did find any type of shielding that would prevent radiation SEUs in
any of the later processors.  I might argue it might be an artifact of an
over-test at the cyclotron at Berkeley, but I wouldn't hang my life on any
Pentium or later chip in the space environment.  Any shielding that is light
enough to be carried into space creates a shower of secondary particles when
hit with a sufficiently energetic cosmic ray, so that most attempts at
shielding make the problem worse instead of better.  Recall also that it
would need to be shielded from all directions, and that to stop a cosmic
particle would take a huge block of lead.

>...The latest chips use 22nm and 32nm technology whereas the 486 used 800
and 1000nm technology. This means that every stray cosmic ray causes much
more damage in a modern chip...  BillK

In space applications, the motto is if it works, don't mess with it.  The
humble i486 works, if you de-clock it from its blazing 40 mega-Hertz and are
old enough to know the definition of the archeo-prefix mega .  If you need
any heavy-duty space-calculations, the way to go is to send the data to the
deck, have the super-computers calculate it and send it back up.


More information about the extropy-chat mailing list