Didn't Steve Jobs...
...get the idea for the iPad after he watched 2001: A Space Odyssey and Star Trek? Or did Judge Koh decide that was unpossible?
Ever dreamed of taking Saturday morning trips to the computer lab at Cambridge University, playing with the equipment, cannibalizing old computers, and building new machines from the bits? We're not talking metal Meccano minnows run on AA batteries, either. We’re talking actual, operating electro-mechanical machines powered by …
When I was at university (in the dark ages, not quite Wilkes era, but not too long afterwards), we had a Control Data CYBER74. It ran exactly this way -- big motor-generator sets in the basement converted line voltage to 400Hz AC -- allowing the power supply components to be smaller was the reason I was given -- which was then used to power the beast.
I believe it was a case of "horses for courses" -- the electronic converters available at the time were more expensive and less reliable than the good old electromechanical way of doing the job.
400Hz is also MIL spec. It's used on ships and planes because 400Hz motors can be smaller and lighter. It's not used on utility grids becaujse long-distance transmission losses are prohibitively high. 400Hz pre-switched-mode power supplies were also lighter, and needed less huge electrolytic capacitors.
Did CDC sell a lot of kit to the US Military? I'm guessing that they did.
"Did CDC sell a lot of kit to the US Military? I'm guessing that they did.
There's a book called "A few good men from Univac," about the founding of CDC. Note 2 things.
Univac definitly did. IIRC they were big suppliers of the US navies standard ISA computers (the YK something or other. Less famous than the USAF's 1750A ISA).
One of those men was Seymour Cray.
The old motor/generator trick is old school power electronics. Right up with the monster mercury arc rectifier (make sure your wearing sunblock and dark glasses if you're standing near one of these when it's running).
400Hz AC -- allowing the power supply components to be smaller was the reason I was given
Indeed. A transformer's power rating is the core area times the working frequency times some factor, so if your frequency goes up, the transformer's size can go down while keeping the same power specification. Also, smoothing capacitors need only buffer 1/800 sec instead of 1/100 (or 1/120) sec., so you can use smaller caps (in value, which translates to physically smaller again).
Actually, no. The Maurice who won the Nobel Prize in 1962 was molecular biologist Maurice Wilkins, not Maurice Wilkes. When Alfred Nobel wrote his will in the 19th century, he didn't foresee that there might someday be computers and thus computer scientists deserving of prizes.
Wilkes lectured on numerical analysis to first-year students - I was one of them. His lectues were wonderfully clear, and tempted me to do the numerical analysis questions in the exams. I found these questions seemed somewhat easier than the questions in the more popular fields.
He was lecturing to a large audience, of maths students plus those doing various sciences. In that large room, he used a microphone, which kept breaking down. Some of the plebs jeered or hissed, others conspicuously walked out. Some seemed to attend the lecture for the sole purpose of walking out when the embarrassment began.
One valuable point he taught was that bit patterns could be interpreted as fractions rather than integers. Years later I had to modify software for Dungeness nuclear power station, where the equations for the control algorithms were implemented using fractions - so for example a temperature might be a fraction of 800 degrees Centigrade. I found it useful to provide two comments for each section of code: first the mathematical equation, and secondly the scaled arithmetic equation. The latter would explain the sundry multiply or divide operations by seemingly arbitrary constants.
There was no floating point arithmetic on small process control computers in those days. I shudder to think how such software would get regulatory approval today.
"As opposed to what exactly? Analogue?"
Yup, many of those. See the tide predictor in Liverpool Museum should you be passing by one day.
http://en.wikipedia.org/wiki/Tide-predicting_machine [ not the Liverpool one]
http://www.liverpoolmuseums.org.uk/wml/collections/physical-sciences/oceanography.aspx
And remember that for a generation before Wilkes', a 'computer' was a retired bank clerk or amenable housewife who would pound away on the Brunsviga according to your instructions.
http://home.vicnet.net.au/~wolff/calculators/Brunsviga/Brunsviga.htm
I worked with the Lab team and Titan in my senior year in Physics, and managed to write a radio-astronomy program that brought Titan to its knees every night. Finally, the Lab took pity on all the experimenters I locked out, and showed me machine loop coding! Fun days, a great group to work with, and they were patient. If I did this today, I'd probably just get a nice message saying "over allocation. Come back next month!"
The result was I decided to be a computer engineer and not a radio-astronomer!