52 | 6 TURING’S GREAT INVENTION
Turing’s plans caught the interest of the British press, with headlines such as ‘ “ACE” may be
fastest brain’, ‘Month’s work in a minute’, and ‘ “ACE” superior to U.S. model’.^8 Behind the scenes,
however, all was not well. Turing’s plan was that Flowers and his assistants from the Colossus
days should build ACE, but this idea soon ran into trouble. Flowers came under tremendous
pressure to devote himself to the rebuilding of Britain’s war-ravaged telephone system, and
despite his willingness to collaborate with Turing, he rapidly became ‘too busy’, he said.^9 The
NPL’s slow-moving bureaucracy seemed unable to make alternative arrangements, and in April
1948 Turing’s manager, John Womersley, reported ruefully that hardware development was
‘probably as far advanced 18 months ago’.^10
While Turing waited for engineering developments to start, he and his assistants spent their
time pioneering computer programming, preparing a large library of routines for the not-yet-
existent computer.^11 Once the hardware was finally working, it was the availability of these
ready-made programs that explained the success of NPL’s scientific computing service.^12 This
took on commissions from government, industry, and the universities, and was the first such
service in the world.
In the end, the pilot model of Turing’s ACE did not run its first program until May 1950
(Fig. 6.2). By that time, a handful of stored-program computers were working successfully, but
Turing’s 1-MHz design left them all in the dust.
Open for business in manchester
Manchester, though, had won the race fair and square in 1948, and Turing moved there shortly
after Baby first came to life.^13 He was tired of NPL’s ineffectual bureaucrats, and itched to get his
hands on an electronic universal Turing machine.
figure 6.2 The pilot model of Turing’s Automatic Computing Engine.
© Crown Copyright and reproduced with permission of the National Physical Laboratory.