Apollo Core Weavers — The Women Who Hand-Wove the Moon Landings

The software that flew Apollo 11 to the Moon was not stored on a disk or a chip. It was threaded by hand through ferrite rings 1.5 mm wide, one wire decision at a time, by women in a Raytheon factory in Waltham, Massachusetts and Navajo women at a Fairchild plant in Shiprock, New Mexico. A wire through the ring was a 1. A wire around it was a 0. The 1975 NASA retrospective on Apollo computing named neither group.

How core rope memory worked

The Apollo Guidance Computer carried two memories. Erasable magnetic core held 2,048 words for in-flight scratch space. Fixed core rope held 36,864 words of program — every line of code that would fly the mission, encoded once and unalterable.

The physics was simple. A read pulse on a sense wire induced a signal in any ferrite ring the wire passed through; rings the wire skirted stayed silent. A program became a three-dimensional routing diagram: tens of thousands of rings, threaded by hand to a wiring chart with no shortcuts and no inference allowed. Each Apollo flight needed six rope modules. A complete set consumed roughly six woman-years of weaving before launch.

A single misthreaded wire invalidated the entire rope. There was no patch — only microsurgical rework or a fresh start. Three to four inspectors signed off every assembly before it moved.

The logic is identical to the tech jacquard loom of 1804: a binary decision (through or around, raised or lowered) repeated across a grid to encode a complex pattern. Ada Lovelace used the loom as her explanatory analogy for the Analytical Engine in 1843. The Apollo weavers were the same idea, scaled to spaceflight.

The Raytheon "LOLs"

Raytheon's Bedford workforce was overwhelmingly women, many recruited from the dying mill towns of eastern Massachusetts where 19th-century textile labor had built the regional economy. Engineers nicknamed them the "Little Old Ladies" — LOLs. The workers were mostly not old. The label did its job anyway: it framed precision aerospace manufacturing as artisanal women's work, payable at artisanal wages.

Margaret Hamilton, who directed the AGC software effort at MIT's Instrumentation Laboratory, carried the title Rope Mother — the senior software lead who shepherded a program from code to woven rope. Her famous contribution to Apollo 11 was the priority-interrupt logic that survived the 1202 alarms during lunar descent. The metaphor in her job title was not decorative. The software team wrote the pattern. The Raytheon weavers were the loom.

Shiprock: indigenous circuits

In 1965 Fairchild Semiconductor opened a plant on the Navajo reservation in Shiprock, New Mexico. At peak it employed more than 1,000 people, the overwhelming majority Navajo women, assembling integrated circuits — some destined for the AGC, most for the broader semiconductor boom.

The job listing called for "painstaking attention to detail, excellent eyesight, high standards of quality." A Fairchild dedication brochure compared IC assembly to traditional Navajo rug weaving and framed the fit as natural. The framing was the wage strategy. Precision technical labor reclassified as indigenous craft pays craft rates. Lisa Nakamura calls this pattern indigenous circuits — and traces its lineage forward to the "nimble fingers" rhetoric used to globalize semiconductor assembly to Southeast Asian women in the 1970s–90s.

The Shiprock plant closed in 1975 after an American Indian Movement occupation over layoffs and working conditions. NASA's retrospective on Apollo computing came out the same year and named no one from the line.

The textile-to-computing thread

Year	Technology	Who threaded it
1804	Jacquard loom punch cards	Women operators in Lyon
1843	Analytical Engine (proposed)	Ada Lovelace, using the loom as the analogy
1950s	Magnetic core RAM	Women threaders, US and Asia
1964–69	Apollo core rope ROM	Raytheon women, Shiprock Navajo women
1970s–90s	IC assembly	Southeast Asian women
2020s	Programmable knitting	Code; the weaver automated out

The constant: information physically encoded into a textile-adjacent structure, performed by women, classified as craft, paid as craft, omitted from the history.

What's contested

The individual stories of Shiprock workers are still being recovered. Most names, shift records, and oral histories are not yet in the public archive — what exists comes from Fairchild's own promotional materials (which had reason to flatter), AIM-era organizing records (which had reason to indict), and a small number of post-2015 oral history projects. The full picture of who built the chips in the AGC is still emerging, and the percentage of AGC integrated circuits that came specifically from Shiprock versus Fairchild's California fabs is not cleanly documented in the open literature.

Hamilton's "Rope Mother" title is sometimes retold as if she personally directed the Raytheon weavers. She did not — Raytheon ran its own floor. The title described her role on the MIT software side of the handoff.

Why this has to do with other realms

The Apollo ropes sit on a 5,000-year line that runs through Andean quipu — knotted cord records that encoded census, tribute, and possibly narrative, and that the Spanish destroyed in part because they could not read them and could not believe a textile was a document. Both quipu and core rope encode information in the topology of fiber. Both were made by indigenous women. Both were dismissed by the credentialed culture around them as decoration or domestic craft until someone noticed the bits. See overview andean textiles.

The structural parallel runs forward too. Modern quantum memory in surface and toric codes encodes a logical qubit in the topology of physical qubits on a lattice — the bit is not in any one location but in the routing. The aesthetic is the same as a core rope, several abstraction layers up. See concept quantum error correction.

An open question

If the Jacquard loom, the core rope, and modern programmable knitting all collapse the distinction between weaving and programming — what changes about the history of computing if you start the timeline in Lyon in 1804 instead of Bletchley Park in 1943, and count the workforce that was there the whole time?

Key sources

Sisters in Making — shorthandstories.com oral history project documenting Raytheon's women weavers; the most complete public record of named individuals.
Lisa Nakamura, "Indigenous Circuits: Navajo Women and the Racialization of Early Electronic Manufacturing" (American Quarterly, 2014) — the load-bearing scholarly account of Shiprock.
David Mindell, Digital Apollo (MIT Press, 2008) — the AGC hardware and software story, light on the manufacturing workforce but rigorous on the rest.
Eldon Hall, Journey to the Moon: The History of the Apollo Guidance Computer (1996) — the AGC chief engineer's account; useful for what it includes and revealing for what it omits.
To verify: the 1975 NASA retrospective ("Computers in Spaceflight: The NASA Experience") and its acknowledgments page — the specific absence is itself the source.
Science News, "The women who hand-wove the Apollo guidance computer's memory" (2022) — recent mainstream account naming Shiprock workers alongside the LOLs.

Abhishek's take

I see the same mispricing when a hand-embroidered yoke sample reaches the floor: the decision looks like design, but the risk sits in the fingers that must repeat it without error. A CAD file can change in an hour; a bad thread path in production burns the season’s calendar. The textbook calls it craft, but on the floor I price it as memory, because the garment carries decisions long before it reaches a store.

Abhishek