Path: utzoo!attcan!uunet!lll-winken!lll-lcc!ames!ncar!oddjob!gargoyle!att!westmark!dave
From: dave@westmark.UUCP (Dave Levenson)
Newsgroups: comp.misc
Subject: Re: Core memory
Message-ID: <246@westmark.UUCP>
Date: 19 Jul 88 03:03:06 GMT
References: <1486@crete.cs.glasgow.ac.uk>
Organization: Westmark, Inc., Warren, NJ, USA
Lines: 29

In article <1486@crete.cs.glasgow.ac.uk>, shoat@cs.glasgow.ac.uk (Dave Shoat) writes:
> I recently took apart a core memory board from a PDP8, just out of
> curiosity. I was quite amazed at the small size of the magnets - I had to
> put the board under a microscope to see any detail at all. The whole thing
> was very intricately made. Does anyone know hoe core memory was manufactured?
> Was it knitted? By hand or by machine? I'd be interested to know.


I don't know how DEC core planes were made, but back in 1970 or so
when I worked for Interdata (more recently known as Perkin Elmer and
then as Concurrent Computer Co) we made core planes by hand.  A
large room in the factory, known as the "core house", was populated
by about 25 employees who "stitched" the wires through the cores by
hand.  They sat at tables with magnifying lamps (the kind where you
look through a 6-inch diameter lens which is surrounded by a
circular fluorescent tube) and watched their work through the glass.

They also made ROMS by hand!  They threaded thousands of wires
through or around individual transformer cores, arranged in a row of
16. Each wire represented an address, and its content was determined
by which cores (or bits) it passed through, and which it passed
around.  When the address line was decoded, it was pulsed, and the
sense wires of the transformers through which the address line
passed sensed this pulse.

-- 
Dave Levenson
Westmark, Inc.		The Man in the Mooney
Warren, NJ USA
{rutgers | att}!westmark!dave