The onboard computers for Apollo—one that flew the command module to the moon and back to earth, and another that flew the lunar module from orbit around the moon to a safe landing, then back up into orbit—were the smallest, fastest, most nimble computers ever created for their era.
Designed and programmed by scientists, engineers and programmers at the Massachusetts Institute of Technology, the computers were marvels of their time—and a view into the computing future. In an era when a small computer was the size of three refrigerators, lined up next to each other, the Apollo flight computer was the size of a briefcase. At a time when computers on earth required punch cards to work, and hours to get results back, the Apollo flight computer had a keyboard and worked instantly. In an era when people using the computers simply submitted their punch cards and waited for the results from computer operators, the astronauts ran the Apollo flight computers themselves.
But in the mid-to-late-1960s, when the Apollo computers were designed, programmed and built, they were just a few years ahead of our ability to manufacture their circuitry. Computer chips and computer memory were in their infancy—indeed, the Apollo computer was the first computer of any significance to use integrated circuits and computer chips.
The Apollo computers were designed with a kind of memory called “core rope memory.” It was the densest computer memory available at that moment in time—between 10 and 100 times more efficient, in terms of weight and space, of any other memory available, absolutely essential on spacecraft where weight and space were always at a premium.
But core rope memory suffered from one small problem: It had to be made by hand.
Each wire representing a 1 or a 0 in the computer program had to be positioned with absolute precision, by a person, using a needle, and wire instead of thread. A wire threaded through the center of a tiny ring-shaped magnet was a one. A wire threaded to the outside of that magnet was a zero.
And so the most remarkable computer of its era—not just a space-age computer, but a spaceship flight computer—had circuitry that was hand-woven, by women, many of them former textile workers, in a Raytheon factory in Waltham, Massachusetts.
The Apollo guidance computer contained just 73 kilobytes of memory—far less computing power than a typical microwave oven today. In all, it contained 589,824 ones and zeros of computer programming—and every single one and zero, every single wire, had to be positioned exactly correctly, or some part of the computer’s sophisticated flight program wouldn’t work right.
Because the women in Waltham weren’t just weaving the memory of the computer, but literally weaving the programming instructions directly—almost all of the Apollo computer’s memory was fixed—and woven by those women. For Apollo, the software was in fact hardware.
It took eight weeks to weave the memory for a single flight computer. The computers in the command module and the lunar module were identical, but their programming was different, and the programs for each Apollo flight were also different.
While tedious, the work demanded attention, skill and experience. Raytheon found that out during Apollo when there was a brief strike in the mid-1960s that included the Waltham factory.
Managers and supervisors attempted to keep the Apollo computer assembly line going by sitting down to do the weaving themselves. According to Ed Blondin, a senior manager at the facility, “Everything they made was scrap.”
History Reads features the work of prominent authors and historians.