- p.37, in Whitehall, explaining the message from space:
- "It's a computer program," he said quietly.
- . . .
- "It's in three sections. ...
The first part is a design -- or rather it's a
mathematical requirement which can be interpreted as a design.
The second part is the program proper, the order code as we call it.
The third and last part is data -- information sent for
the machine to work on."
- And p.39:
- "Our newest computers still work in microseconds.
This is a machine that must operate in
milli-microseconds [10-9s], ..."(1)
- p.54-55, Flemming telling Judy about the computer:
- "The memory is in the core and the core is held in a
total vacuum to within a degree or two of absolute zero."
- "Each core is built up of alternating layers ... half a thou. thick ...
a complete yes-no gate circuit(2)
on a spot of metal you can hardly see."
- "Is that the equivalent of a brain cell?"
- "If you like."
- . . .
- "The core's a three-metre cube.
That makes several millions of millions.
And there are six cores."(3)
- "It's bigger than a human brain."
- "Oh yes. Much bigger. And faster. And more efficient."
- (1)In the 1960s a respectable super computer,
with its own building and 24-hour staff, might have
a speed of 106 instructions per second, as suggested,
and 106 bytes of main memory.
Giga-hertz clock speeds (109) became commonplace c2000
with, for example, the Intel(!)
- (2)Does that maybe sound more like a
processing gate rather than just a memory bit?
- (3)So let's say
about 4×1013 brain cell equivalents.
For 1961 that is a good margin over a human in terms of neuron numbers.
But if the contest were in terms of synapses,
it is now known to be much tighter:
- There are "only" about 2×1010 neurons in a human brain.
However, each neuron is connected to several thousand others on average,
perhaps 1014+ synapses in total
-- far higher connectivity than in electronic circuits (as of 2012).
The A for A text does say "cells", that is neurons.
- See Do we have brain to spare?, D. A. Drachman,
Neurology 64(12) 2005