Path: utzoo!utgpu!water!watmath!clyde!att!whuts!mhuxh!mhuxu!mhuxt!mhuxi!mhuhk!mhuxo!ulysses!thumper!karn
From: karn@thumper.bellcore.com (Phil R. Karn)
Newsgroups: comp.misc
Subject: Re: Core memory
Summary: use of core memory
Message-ID: <1250@thumper.bellcore.com>
Date: 21 Jul 88 18:58:35 GMT
References: <1486@crete.cs.glasgow.ac.uk> <1010@garth.UUCP> <3397@phri.UUCP>
Organization: Bell Communications Research, Inc
Lines: 24

> 	Is it?  I remember reading somewhere (IEEE Spectrum a few years
> ago?) that the space shuttle uses core memory because it is 1) radiation
> hard and 2) static (i.e. no loss of memory on loss of power).  Apparently
> the slow speed, power-hungry operation, and low information density are
> outweighted by the other factors.  Can anybody corroborate this?

Yes, the Shuttle computers use core memory, but these two reasons are a
bit rationalized. The real reason is that the Shuttle's computers (IBM
AP-101s) were already obsolete by the time the Shuttle project chose
them.  Semiconductor RAM technology is now quite capable of meeting the
needs of spaceborne computers; AMSAT-Oscar-13 is flying 32K bytes of
error corrected, radiation hardened CMOS RAM from Harris Semiconductor.
Even this technology (each physical chip is only 4K bits) is way behind
the terrestrial state of the art because of the special need for
radiation hardening (100 kilorads in our case). But even conventional
dynamic RAM memories can withstand 2-3 kilorads, which is 4-6 times the
typical human lethal dose.

How the space program ever got the public image of being on the leading
edge of computer technology is a mystery to me. The process of designing
and certifying a computer system for space flight guarantees that it
will be obsolete by the time it flies.

Phil