Definition
Random-access memory that stores binary bits in tiny magnetic cores whose magnetization can be reversed, using wire windings for selection, reading, and writing.
What it is
Magnetic-core memory consists of many small ferrite cores whose magnetization direction can be reversed. Intersecting row and column wires select a target cell. After its magnetic state is read, the bit destroyed by the read is written back, forming random-access main memory.
Why it emerged and what problems it solved
Delay lines depend on the position of data in the medium, making them less flexible for random access to programs and data. Magnetic-core arrays provided more consistent addressed access and the ability to retain data without power, allowing stored-program computers to grow in scale.
Why it matters and what it enabled
Magnetic cores placed main-memory capacity, speed, and reliability at the center of system design and became a benchmark for comparing the cost and density of semiconductor memory. Large-scale replacement of cores began only after MOS DRAM reached sufficient density and sufficiently low prices.
Principles and impact
Problems solved
- Addressed the serial-access limitations of delay-line memory.
- Provided stable, scalable random-access main memory for large stored-program computers.
Principles
- The two remanent magnetization directions of a core represent 0 and 1.
- Currents in row and column wires select the target core; reading usually destroys the state and requires rewriting.
Improvements
- Made random-access latency more predictable and retained magnetic states after power was removed.
- An array structure made it easier to expand capacity as machines grew.
Significance
- Served as computer main memory for an extended period from the 1950s through the 1970s.
- Provided performance and cost benchmarks for semiconductor RAM replacing main memory.
History and applications
People
An Wang、Jay Forrester
Organizations
Harvard University、MIT Lincoln Laboratory
Events
Magnetic-core-memory patents and prototypes in 1949、Became a computer main-memory solution in the 1950s
Applications
Mainframe main memory、Real-time control computers、Early minicomputers and spacecraft computers
Limitations
- Manufacturing required extensive manual wire threading, resulting in relatively large size and high cost.
- Read/write wires and driver circuits are complex, and magnetic materials constrain speed and density.
Upstream and downstream relations
Upstream
Downstream
Sources: 1970: MOS Dynamic RAM Competes with Magnetic Core Memory on Price