More Secrets from the U.K. Revealed
February 3rd, 2010 by Paul Ceruzzi
You would think that after all these years we would know pretty much the whole story of British code-breaking computers during World War II, but we don’t. Here is a link to a BBC series, in which Simon Lavington discusses “Oedipus”–a post-war proto-supercomputer, the details of which are only now beginning to emerge. See my earlier post on Sam Snyder; also my post on the British lead in computers and how they lost it to the U.S.
Lavington says that one of the unique features of the machine was its use of associative memory. This concept was apparently independently discovered at least three times, by the British, by IBM, and lesser-known, in 1943 by Konrad Zuse (who has a sketch of associative-memory addressing circuits in his autobiography, Chapter 5). I’ll bet Zuse was the first.
I understand the need for secrecy, but what if this remarkable technology had not been kept under wraps? Now, with Google, the genie is out of the bottle anyway.
—
February 6th, 2010 at 1:41 pm
According to Rex Seeber (co-designer of the SSEC among other things) when he first came to work for Howard Aiken on the Mark I (circa 1944), Aiken asked him how he thought the machine’s memory worked and Seeber speculated on how it could work on an associative principle like human memory before Aiken explained that was not how the machine worked. Seeber worked on an IBM implementation of associative memory in the 1960s so the idea seems to have caught his imagination.
So perhaps the mystery is why it took so long for associative memories to be built?