COPElAND | 155
A Colossal computer
Telephone engineer Tommy Flowers (Fig. 6.1) knew that he could build something better than
Heath Robinson. In the years before the war he had begun to pioneer large-scale digital elec-
tronics. Few outside his small group at the Post Office Research Station at Dollis Hill in North
London knew that electronic valves could be used reliably in large numbers.
According to conventional wisdom, valves were just too flaky to be used en masse: each
one contained a hot glowing filament, and the delicacy of this filament meant that valves were
prone to sudden death. The more valves there were in the equipment, the greater the chance
that one or two would fizzle out in the middle of a job. Plagued by continually blowing valves, a
big installation would be completely impractical, or so most engineers believed: it was a belief
based on experience with radio receivers and the like, which were switched on and off fre-
quently. But Flowers had discovered that so long as valves were left running continuously, they
were very reliable. He also discovered ways of boosting the reliability even further, such as using
lower than normal electric currents. As early as 1934, he had successfully wired together an
experimental installation containing 3000 to 4000 valves. The equipment was for controlling
connections between telephone exchanges by means of tones, like today’s touch-tones.
Digital electronics was a little-known field in those days. In 1977 Flowers told me that, at the
outbreak of war with Germany, he was possibly the only person in Britain who realized that
valves could be used for large-scale high-speed digital computing.^11 So he turned out to be the
right man in the right place when he was sent to assist the codebreakers at Bletchley Park. He
realized that Heath Robinson could not give them the accuracy and speed they needed. Heath
Robinson contained no more than a few dozen valves—a puny number as far as Flowers was
concerned. He knew that by using valves in large quantities he could store information that
Heath Robinson could store only on wear-prone paper tape. Flowers proposed an electronic
monster containing approximately 2000 valves.
Newman took advice on Flowers’ proposal, consulting other specialists in electronics, and
was told that such a large installation of valves would never work reliably. In consequence,
Bletchley Park declined to support it. But in his laboratory at Dollis Hill Flowers quietly got
on with building the all-electronic machine that he knew the codebreakers needed. He and his
team of engineers and wiremen worked day and night for ten months to build Colossus, as the
giant computer was called—they worked until their ‘eyes dropped out’, Flowers recalled. All
this was ‘without the concurrence of BP’ (Bletchley Park), he said, and was done ‘in the face
of scepticism’. ‘BP weren’t interested until they saw it [Colossus] working’, he recalled with wry
amusement. Fortunately his boss, Gordon Radley, had greater faith in Flowers and his ideas
than Bletchley Park did, and gave him whatever he needed.
On 18 January 1944, Flowers’ lads took Colossus to Bletchley Park on the back of a lorry. It
was a significant moment in the history of computing—the delivery of the world’s first large-
scale electronic computer. The computer’s arrival caused quite a stir. ‘I don’t think they under-
stood very clearly what I was proposing until they actually had the machine’, Flowers said. ‘They
just couldn’t believe it!’ The codebreakers were astonished by the speed of Colossus—and also
by the fact that, unlike Heath Robinson, it would always produce the same result if set the same
problem again.
The name ‘Colossus’ was apt—the computer weighed about a ton (Fig. 14.8). Flowers’ gigan-
tic computer began active service against the German messages on 5 February 1944. As Flowers
noted laconically in his diary, ‘Colossus did its first job. Car broke down on way home’.