Photo: ArnoldReinhold · CC BY-SA 4.0
~4 KB
The computer that guided Apollo astronauts to the Moon had about 4 kilobytes of working memory.
5 min readLast updated
The short answer
Yes. The Apollo Guidance Computer had 2,048 words of erasable memory, roughly 4 kilobytes, and about 36,864 words, roughly 72 kilobytes, of fixed read only memory. With that, it helped steer astronauts to the Moon and back.
Why it's true
Each Apollo mission carried two guidance computers, one in the command module and one in the lunar module. They were designed at the MIT Instrumentation Laboratory and built by Raytheon. Astronauts used a keypad and display called the DSKY to enter commands as pairs of verbs and nouns.
Memory was precious. The erasable memory, used for working data, held 2,048 words of 16 bits, around 4 kilobytes. The fixed memory, which stored the programs, was made using core rope, a technique in which wires were woven through or around tiny magnetic rings by hand. A wire passing through a ring meant a one, and a wire going around it meant a zero.
The computer ran at about 1 megahertz and could perform tens of thousands of operations per second. That sounds tiny today, but it was among the most advanced computers of its time, and it was one of the first to use integrated circuits on a large scale.
By the numbers
2,048 words
Erasable memory, about 4 KB
36,864 words
Fixed memory, about 72 KB
~1 MHz
Approximate clock speed
1969
Year it helped land Apollo 11
How we know
NASA gave MIT the contract to develop Apollo's guidance system in 1961, soon after President Kennedy set the goal of landing on the Moon.
Software for the computer was written by a team led by Margaret Hamilton, who helped popularize the term software engineering. Her team designed the software to prioritise the most important tasks if the computer became overloaded.
That design proved vital during the Apollo 11 landing in July 1969, when the computer raised a series of 1202 and 1201 alarms. The software dropped less important tasks and kept the landing going, and Neil Armstrong and Buzz Aldrin landed safely.
Hamilton received the Presidential Medal of Freedom in 2016 for her work on Apollo software.
Small memory, giant leap
Programrs worked hard to fit essential calculations into tiny amounts of memory, reusing every word they could.
The computer also had to be extremely reliable, because a failure far from Earth could be deadly.
Today, enthusiasts have built working replicas and simulators of the Apollo Guidance Computer, and its source code has been published online.
Its story is a reminder that clever design can matter more than raw computing power.
Testing the software was as important as writing it. The MIT team used simulators to run through landings and emergencies again and again, checking how the programs behaved when things went wrong. Hamilton's young daughter once crashed a simulation by pressing keys during a pretend flight, which led her to think about how astronauts might make similar mistakes and how the software should handle them.
Later, on Apollo 14, a faulty abort switch threatened the landing. Engineers on Earth quickly wrote a workaround that astronauts typed into the computer by hand, letting the mission continue to a successful landing.
Worth remembering
- The Apollo computer had about 4 KB of working memory.
- Its programs were woven into core rope memory.
- Margaret Hamilton's team wrote its software.
In context
A simple calculator or a digital watch has more processing power than the computer that helped land people on the Moon. What made Apollo succeed was not raw power but careful design, clever software and skilled people.
A closer look
Weaving memory by hand
Core rope memory was assembled by skilled workers, many of them women from textile and watch industries, who threaded wires through tiny cores with great precision.
Engineers nicknamed it LOL memory, for little old ladies, a reflection of the workforce at the time.
Software that saved the landing
The 1202 and 1201 alarms during Apollo 11 happened because a radar switch setting sent extra data to the computer, overloading it.
Hamilton's priority based design let it restart and keep the most important landing calculations running.
Integrated circuits in space
The Apollo program was one of the biggest early buyers of integrated circuits, helping drive down their cost.
That boost helped lay the groundwork for the microchips in today's computers and phones.
Related facts
The computer helped put those footprints there. Footprints on the Moon.
Apollo was an early major user of microchips. Integrated circuits.
The Intel 4004 came soon after Apollo. An early microprocessor.
Computers shrank dramatically in just two decades. ENIAC's huge size.
Key terms
- DSKY
- The display and keyboard astronauts used to talk to the Apollo computer.
- Core rope memory
- Woven wire memory that stored Apollo's programs.
- Erasable memory
- Memory that could be changed during a mission.
Myth vs reality
Myth
The Apollo computer was primitive and barely worked.
Reality
It was a carefully designed, reliable machine that was cutting edge for its time.
People also ask
How much memory did it have?
About 4 kilobytes of erasable memory and about 72 kilobytes of fixed memory.
What was core rope memory?
Program memory made by weaving wires through or around magnetic cores, storing ones and zeros permanently.
Who wrote the software?
A team at the MIT Instrumentation Laboratory, led by Margaret Hamilton.
What were the 1202 alarms?
Warnings that the computer was overloaded. The software handled them by dropping less important tasks.
How did astronauts use it?
Through a keypad and display called the DSKY, entering verb and noun codes.