As I was born in 1969, the year that man first set foot on the moon I of course have no recollection of this historic event but also, keeping track of how long ago this happened is obviously a no brainer for me. In todays world of tech and social media most users are familiar with many acronyms used as a shorthand tool in which LOL means 'Laughing Out Loud'. I am sure that they had lots of fun aboard Apollo 11 but back in the 1960's LOL mean't something entirely different. It stood for 'Little Old Ladies' and they had a crucial role to play in this mission. It is a pity we don't hear more about these unsung hero's well I never did until my other half mentioned them. I thank him for inspiring this blog post :) The LOL's worked for Raytheon (which co-incidentally my better did too once upon a time). Intrigued? Then read on, dear reader, read on! The computer for this mission was developed by MIT and looked like this:
 |
Apollo Guidance Computer (AGC)
|
|
|
Initially it was thought that the AGC would only require 4KB of ROM (Read Only Memory) and 0.5KB of erasable random access memory (RAM) would be adequate. Perhaps it is just my humble opinion or the benefit of hindsight or something but that seems a bit naive to me. I am no computer whizz though so sorry if I'm wrong about that. As the software grew, the need for more ROM to store it on increased, so the final memory specs were 36KB ROM and 2KB of erasable memory. In those days, computer programmes were not stored on a hard disk in the modern style but in ROM which was fabricated by weaving a copper wire either through or around a tiny magnetic core. If a wire passed through a core it represented a '1' and around it a '0'. In this way, the software was painstakingly woven together by a team of women weavers at a factory. The team at MIT called it the 'LOL method' who threaded the cores in the Raytheon factory, core by core, wire by wire, bit by bit, the software programmes were woven into the hardware. Although if you ask me, they don't exactly look like old ladies to me.
The end result was tough and space flight enviroment-resistant, and could be wrapped carefully into a relatively tiny space on board the space craft. The prototype designs looked like tangled pieces of rope so the team dubbed them 'rope memory'. It took approximately six weeks to manufacture a rope and so all programming had to be stopped six months prior to a mission, to allow time for manufacture and then testing before flight. Once the rope module had been produced alterations to the code were impossible. The computer's 2KB RAM needed to write and read live mission data and was manufactured from cores in this way with the '1's and '0's written temporarily to each core by using a high enough current to flip the cores magnetic direction between clockwise and counter-clockwise, and read using a second lower-powered pulse of current to sense this magnetic field direction. To make the most of this very limited erasable memory, different programmes running at different times used the same blocks of memory in a sort of 'time-share' way, overwriting each other as they ran.
 |
| close up of fine wiring achieved |
 |
| core rope memory |
 |
| a finished core rope memory |
|
|
|
|
(Credit for images:Raytheon)
This computer was more akin to the modern computer system in a car as opposed to a personal desktop or laptop computer with a QWERTY keyboard. I will leave you with one further piece of info which I found fascinating but which also amused my highly developed inner child. Back then the unit of memory in the AGC was the 'word' (byte was never used). Sounds cool doesn't it? ;) The 'word' was made up of 15 bits for memory storage and 1 extra bit for error detection code called 'parity'. Interesting? What do you think?
So thanks to the LOL's of Raytheon and to the Haynes Apollo 11 manual
No comments:
Post a Comment