Skip to main content

The National MagLab is funded by the National Science Foundation and the State of Florida.

Magnetic Core Memory – 1949

At the dawn of the computer age, magnetic core memory helped make data storage possible, and showed surprising staying power in a field where components are constantly being replaced by new and improved products.


Magnetic Core Memory – 1949

In the rapidly evolving world of computers, critical components often follow a path from cutting-edge to obsolete in short order. There are occasions, however, when a key mechanism displays prolonged staying power.

One such example is magnetic core memory. As one of the first stages of random access computer memory, it features tiny, circular ceramic structures (known as cores), all of which contain data and are linked by a network of wires. This storage system is commonly referred to as "core."

Core memory owes its existence to two American physicists who emigrated from China. Way-Dong Woo and An Wang are credited with developing the pulse transfer controlling device in 1949. The name is a reference to the means by which the electromagnetic influence of the cores regulates the transfer of electricity in electromechanical structures.

Within two years of the work by Wang and Woo, two significant advancements accelerated the path to the modern computer. First, Wang developed the write after read cycle, a program that kept the words from being erased at the same time as they were being read. Next, was the establishment of the coincident-current system, which allowed massive groupings of cores to be operated by far fewer wires than had been possible previously. The system necessitated the placement of a wire at a 45-degree angle to the structures, a task that simply could not be done with a machine at the time. As a result, workers with superior hand-eye coordination, using microscopes, did the job.

Within a decade, industrial plants in Asia were manufacturing core. With low wages and low overhead, the price of core plummeted to a level that made it the universal choice for memory, replacing other more costly systems.

As core memory evolved, its costs continued to fall, from about a dollar per bit, all the way down to just $.01 per bit. Even though core had already been in use for several years by this time, Wang did not receive a patent until the mid 1950s. Following a flurry of lawsuits, International Business Machines (IBM) settled with Wang for several million dollars and took complete ownership of the patent. He used the windfall to expand Wang Laboratories.

At its peak, core memory was one of the most highly regarded segments of a group of similar technologies, which utilized the magnetic tendencies of various resources to perform various computer related functions. At the same time, vacuum-tube electronics had reached an advanced stage, but did not last long and required too much electricity. On the other hand, magnetic devices contained many of the best features of the solid-state components that would eventually take their place.

Because it did not require electricity to maintain its contents, core memory had the advantage of being less expensive and easier to store. Additionally, it is impervious to the effects of radiation, a significant consideration given its use in military equipment and in automobiles. The first generation of computers aboard the Space Shuttle used core memory, which demonstrated its durability by remaining workable even during the catastrophic explosion of the Challenger in 1986. Nevertheless, core memory continued to fade from the high-tech field, gradually being replaced by semiconductors.