For The Mother: Apollo Computers

The Display and Keyboard (DSKY) of an Apollo spacecraft. (prepared by the Wichita State University Media Services)

The Mother asked: I took kid #4 to NASA today (you can do stuff like that when you're home schooling). We were at old mission control where the guy explained that the computers they used to control the Apollo missions had 256K RAM each, and each one was the size of the room we were in. So 13 yo had just watched Apollo 13, and they said they had a computer in the command module, and they shunted its functions to the one in the LEM during their flight. My son wanted to know how they managed to package computers in the spaceships with their limited room when they filled basements here on earth.

This is an excellent question because, among other things, it helps show how space travel has had a tremendous impact on the rest of us. You were not wrong about computers and I'll tell you, that eventually I believe the computers we have today were all but inevitable; however, I would not be surprised if they had been much slower in coming if we hadn't needed computer power for spaceflight. Here's why (and the history of computers in US human spaceflight):

The transition from core transistor logic to integrated circuits. For those of you whose eyes just rolled back in your heads, you might want to skip the rest. Core transistor logic was the standard at the time, but it was big and bulky (relatively speaking) and had some disadvantages as quoted from here:

• It could not be complemented, a very important basic operation in computer arithmetic that changes a one to a zero or vice versa.
• It had the characteristic of "destructive readout," in which a datum read from a flip-flop using core transistor logic loses the datum; that forces the inclusion of a circuit to rewrite the datum if it is to be retained after the read cycle.
• Memory cycle time could not be fixed: in Block I it was an average of 19.5 milliseconds, which was quite slow for computers at the time, and the varying cycle caused timing problems within the machine²⁹.

Big deal, you say. We use ICs for everything! Yes, we do, couldn't do what we do today without them, but, at the time, they were just three years old and NASA's always been leery of infant technology. Actually, everyone was. But MIT was building the computer and they wanted to use ICs - and had to convince NASA to let them. ICs were so infant, in fact, that, in 1963, the year after MIT decided to use them, 60% of the total US output of microcircuits were used for the Apollo prototype construction. That meant that the Apollo computer ended up with 36 K with 2 K erasable memory. That limited memory was an issue but they got computer done in under 70 pounds and 70 W.

So, there you go. Have fun.