A simulator needed memory that could keep pace
MIT’s Whirlwind project was meant to calculate an aircraft’s changing position and instruments fast enough for a pilot to interact with the simulation in real time. That demanded more than a computer that could eventually produce the right answer: it had to hold the latest data and retrieve it quickly, again and again.
Early electronic storage methods were often unreliable or awkward for this job. Jay Forrester’s answer was to store each bit—the computer’s basic unit of information—in a tiny ring of magnetic material. The rings could be packed into a grid, with wires running through them.
Selecting one ring with wires
A current sent through a particular pair of wires selected the ring where they crossed. The current’s direction set the ring’s magnetization one way or the other, representing a 0 or 1. Once magnetized, the ring stayed in that state even when power was off. That made the stored information persistent without requiring electricity to hold it in place.
To read a bit, the computer tried to switch the selected ring and detected whether its magnetic state changed. In early core-memory systems, that read could erase the bit, so the computer had to write it back afterward. The tradeoff was worthwhile: the system could access a specific bit quickly and reliably, rather than waiting for information to move through a long sequence.
From flight calculations to general-purpose computers
Core memory gave Whirlwind a practical way to keep the data and instructions needed during a live simulation. The solution also proved useful beyond that project: magnetic-core memory became a widely used form of computer storage for decades, before semiconductor memory displaced it.
The striking link is that a demanding flight simulator helped drive a change in how computers remembered. Whirlwind needed to respond to a pilot without losing the thread of its calculations; Forrester’s rings let a computer keep that thread even through a power interruption.
