Difference between revisions of "Fabritek Core Memory"
m (cat caps) |
m (De-dead-end page) |
||
| Line 1: | Line 1: | ||
[[Image:Moby.JPG|220px|right|thumb|The "Moby" Memory]] | [[Image:Moby.JPG|220px|right|thumb|The "Moby" Memory]] | ||
| − | 256K x 40 bit core memory. | + | The '''Fabritek Core Memory''' was 256K x 40 bit [[core memory]] for the [[PDP-6]] and [[KA10]]. |
Former Fabri-Tek engineer Gary Andersen told this story: | Former Fabri-Tek engineer Gary Andersen told this story: | ||
Revision as of 01:37, 16 January 2024
The Fabritek Core Memory was 256K x 40 bit core memory for the PDP-6 and KA10.
Former Fabri-Tek engineer Gary Andersen told this story: It was one of the first 1 MB add on memories, if not the first, built by a third party vendor. I gave a paper on the design at the Spring Joint Computer show in Manhattan in May of 1965. Dr. Marvin Minsky, from his artificial intelligence lab, [was in the audience]. Unshaven, in a white dress shirt hanging out in back, and in sock-less sandal shoes he came up to me and said, “I want one.”
I said, all we have is a prototype. His response was, “I don’t care. Thanks. When you be in Boston again?” Two weeks later our sales guy and I negotiated a 1 MB add on for his DEC PDP6 computers, one of only 26 PDP6 computers made, for about $230,000 for a 5 microsecond cycle time memory that used 80m ferrite cores, plus another $5000 for every 1/2 microsecond better cycle time, I believe. He wrote the order number on the back of an envelope.
We didn’t believe it, but a week later a hard copy purchase order came from MIT. It was a real chore. Not only the memory, but also the interface. As I remember DEC wasn’t real cooperative. It was a long haul, but we finished and got paid $240,000, I believe
