History of Computing and the Space Race

In the 1950s and 1960s, the US and USSR were locked in a quest to go to the Moon commonly referred to as the "Space Race." The Space Race famously started with the USSR launching Sputnik (the first man-made satellite launched into orbit), and immediately the United States entered a period of rapid scientific development. While many initially think of the scientific advancements obtained from the ensuing missions at NASA as physics-, geology-, and mechanically-focused, arguably among the most important developments made were in computing. In fact, were it not for computing the Apollo program would not have been able to successfully make it to the Moon.

The physics behind rocketry are very unforgiving. To get to Low Earth Orbit, you can only have a maximum of 4% of a rocket’s mass be payload (non-rocket-related mass such as the Astronauts in the rocket or their space capsule)—the remainder of the mass must be disposed of along the way or used up as fuel to propel the rocket upward. This means that for every pound of payload the rocket has to use over one hundred pounds of fuel to leave the atmosphere. Unfortunately, going to the Moon is even more difficult than entering Low Earth Orbit since you have to go there AND return back. This ultimately meant that only a measly 1% of the Apollo rocket’s mass would be able to be payload.

This created a great difficulty for the scientists working on the program. They had a couple options. The first option (and arguably the most straightforward) would be to make a massive rocket that would go to the Moon, land vertically on the moon, and then fly back. The sheer size and weight of this option made it simply outrageous to implement, so it was quickly dropped. The second option was to send all of the parts for the spaceship that would go to the Moon into Low Earth Orbit piecemeal and to assemble them in orbit. That would allow NASA to bypass the restrictions put upon the mass of the payload of the rocket imposed by escaping Low Earth Orbit (which is by far the most taxing restriction on a rocket’s mass budget). However, this option would cost an obscene amount of money in rocket launches and would be completed around the year 2000. Since the Apollo Program and the Space Race were President Kennedy’s pet projects, he wanted the program done sooner than that so that option was also eliminated.

The third possible way to go to the Moon was a relatively fringe concept known as "Lunar Orbit Rendezvous." To summarize, the Apollo rocket would leave Earth, detach a small space capsule after escaping Low Earth Orbit which would fly to the Moon, detach an even smaller piece of that capsule to land on the Moon as a Lunar Lander, relaunch part of that Lunar Lander from the Moon’s surface and reattach it to the capsule in Lunar Orbit, and (once fully disembarked by the astronauts) then detach the part of the Lunar Lander just reattached to the capsule, and then lastly fly back to Earth in the capsule. The key advantage of this idea is that it allowed for a much larger payload since the mass of the rocket being propelled would be reduced significantly over the course of the mission. As the rocket size shrinks (old tanks and pieces of the rocket are ejected), the amount of non-rocketry related mass decreases as well. This allows a smaller amount of fuel to go a bit further since it is not propelling as much deadweight. However, there was one massive draw to this concept: coordinating the two spacecraft in Lunar orbit would be incredibly difficult and complicated. Furthermore, due to the shapes of the spacecraft being linked together, the pilot of each of the crafts would be unable to see the other craft while navigating the rendezvous. Nevertheless, since this option presented NASA with a great balance of time and payload size, NASA chose this concept and developed the Saturn V rocket.

Since rendezvousing the two orbiting spacecraft by hand would be incredibly difficult, NASA needed an alternative approach. As a result, the first contract awarded for the Apollo program was awarded to MIT to develop the Apollo Guidance Computer (AGC). This 16-bit, 2mHz computer with a custom Assembly language was one of the first integrated circuit computers (Wikipedia). The computer ran a custom-made real-time OS and navigation programming made on a team led by Margaret Hamilton and took approximately 1400 man-years of effort to produce.

Since the navigational and maneuvering programs could be damaged by the various conditions encountered in spaceflight on the prevalent (at the time) magnetic-core memory, MIT came up with a new and unique type of read-only memory: core rope memory. In core rope memory, programs were stored via physical wires going around (equivalent to a "zero") and through (equivalent to a "one") various peg-like cores on a board. Sewing these core rope memory based programs was incredibly difficult, so the best seamstresses from around the country were brought in due to the fine precision to which they could work with their hands.

From Apollo 4 to Apollo 8, the various conventional capabilities of the AGC as typical flight computer were tested. With Apollo 8, it was established that the computer could successfully operate as a guidance computer from liftoff to around the moon and back. Finally with Apollo 9 the ability of the AGC to successfully autonomously complete the rendezvous was tested. Since the mission was manned it would be too dangerous to test it for the first time around the Moon, so they tested it in Low Earth Orbit. This mission went perfectly, so then with Apollo 10 NASA tested the rendezvous maneuver in Lunar Orbit. This worked as well. As a result, NASA was confident that they could complete a Lunar Orbit Rendezvous on the way back from successfully landing on the Moon and successfully send men to the moon and bring them back alive. Two months after Apollo 10, Neil Armstrong and Buzz Aldrin successfully set foot on the moon, planted the American Flag, picked up some rocks, and returned home in Apollo XI. Were it not for the computers that allowed the various components of the Apollo missions to rendezvous successfully, the United States would not have been able to reach the Moon and return to Earth as quickly, as safely, and as successfully as it did in the late 1960s and early 1970s.