From: John Conover <john@email.johncon.com>
Subject: Re: Computing As She Really Is. Was: Re: Life-Advancing Work of Timothy Berners-Lee
Date: 13 Nov 1999 19:22:42 -0000
David Lloyd-Jones writes:
>
> Gene Wirchenko <genew@shuswap.net> wrote
> > Brian Inglis <Brian.dot.Inglis@SystematicSw.ab.ca> wrote:
> > > "David Lloyd-Jones" <icomm5@netcom.ca> wrote:
> > [snip]
> > >256K *core* -- RAM is a PC term.
> > ^^^
> > Odd how I heard it before PCs.
>
> I think you're both right -- and me too. :-)
>
> "Core" started to seem old fashioned, and then ridiculous, around then as
> semiconductor memory came on line. Remember it wasn't core as being in the
> core of the machine, but rather this kludge made out of iron magnetic cores
> strung together with, ugh, wires.
>
I think maybe that RAM was introduced by Alonso Church, in his
hierarchy of machines, (state machine, stack machine, random access
machine,) thesis in 1936, which was the formalization of the theory of
algorithms that we use today to program computers.
The ENIAC/EDVAC of the 40's used mercury delay lines, (ie., a tube
filled with mercury, with a piezo-electric transducer on each end,
where the data was maintained as constantly refreshed sound waves-kind
of like a shift register,) which gave way to the use of long
persistence phosphors in CRTs in the early 50's as the memory element
in the RAM architecture. And then on to magnetic cores in the late
50's, and semiconductor memory in the late 60's.
Memory technology, and its trend in decreasing costs, has been the
"driver" of the ubiquity of computational machines-about 70% of the
world semiconductor business is RAMs. (The transistor count of the
memory elements-about 4 transistors per bit-in a modern PC dwarfs the
transistor count in the rest of the entire system; the DRAM in a PC
accounts for about 2 billion transistors!)
John
--
John Conover, john@email.johncon.com, http://www.johncon.com/