20 Oct
1997
20 Oct
'97
10:59 p.m.
Philip.Belben(a)powertech.co.uk wrote :
Are you _sure_ the ABC was electronic? I was under
the impression that
the 1930s machines were all (almost all?) relay computers. Konrad Zuse
is the pioneer whose name is oft mentioned here...
As I see it, the sequence of events is as follows:
The 1940s saw the valve (vacuum tube) computers begin to emerge - in
some order (still in debate) ENIAC, Univac and the hush-hush British
project, Colossus (hush-hush because it was part of the war effort), all
appeared in 1943 I think. Colossus currently claims to have been first,
but it is hard to verify with all the wartime secrecy that surrounded
it.
All the "computers" prior to the Manchester machines did not store their
programs in memory, they are more accurately termed sequence controlled
calculators.
The Univac came later, it was actually first delivered in 1951, just a
month or so after Ferranti delivered the first commercial computer, the
Mark I to Manchester University in February 1951 - and no I don't
remember that particular event personally ;-).
The Manchester machine claims to be the first _stored
program_ machine.
It was the first electronic computer, and I think also the first
computer, to hold its software in main memory. It was far too small for
this to be sensible - the purpose was to demonstrate the principal of it
with a view to using similar hardware and software designs on larger
machines in the future. All previous machines had a main store for
data, and a programming panel for patch leads etc. to hold instructions.
Again yes, the Manchester machine had a main memory of 32 words each of
32 bits, it had one accumulator and an instruction set of only 6
instructions. It was built in 1948 to prove the reliability of the
Williams tube storage system.
The very same hardware was then extensively developed into a fully
fledged computer over the next two years. Ferranti, under a government
contract, built 6 or 7 production versions, the first of which went to
Manchester as stated above.
Soon after the Manchester machine ran, the EDSAC
project in Cambridge
pulled ahead with a large scale stored program (Von Neumann) machine.
Professor Wilkes, who ran this project, said in one talk he gave that
they had wanted to include floating point in EDASC, since it was already
available on many relay computers, but this had to wait until a later
design...
Indeed, the interests of the Manchester team were in the hardware
design, the Cambridge team were more interested in the programming and
uses of computers.
So while Manchester developed hardware technologies, Cambridge developed
software.
Manchester has a long and illustrious history of firsts in the computing
area. As well as the first working stored program electronic computer,
they were the first to incorporate index registers, the first working
transistor computer, and the first virtual memory.
That is not to say that others were not working on the same or similar
lines, its just that Manchester managed to get there first, sometimes by
just a month or two.
More info about the Manchester machines and the rebuild project at
<http://www.cs.man.ac.uk/CCS/ssem/ssemhome.htm>
--
Hans B. Pufal : <mailto:hansp@digiweb.com>
Comprehensive Computer Catalogue : <http://www.digiweb.com/~hansp/ccc/>
_-_-__-___--_-____-_--_-_-____--_---_-_---_--__--_--_--____---_--_--__--_