What is core memory?
Core memory was a common form of random access memory (RAM) from the mid-1950s to the mid-1970s. It was developed at MIT in 1951. The memory used magnetic rings called kernels through which wires passed to select and sense the contents of the kernels. With the advent of memory based on semiconductor technology, core memory became obsolete, although some still refer to a computer's main memory as core memory.
The core memory is also referred to as magnetic core memory.
The function of the core memory was based on the Hysteresis the magnetic material used to make the rings. Each core in core memory was used to store one bit of information. The cores can be magnetized clockwise and counterclockwise. The value stored in the core depended on the direction of magnetization. Access to the core memory comprised read and write cycles.
The read cycle would cause the memory contents to be lost, while the write cycle would restore the contents of the memory location. A read cycle must be followed by a write cycle consequences. Another salient feature of core memory is its non-volatility, which means that its contents are not lost once the power supply is interrupted. Special logic was included in the memory controller to ensure that the memory contents were not changed if the power supplies were not showing their normal values.
Non-volatility was one of the greatest advantages of having core memory in the early years of memory development.
The core memory was quite slow and initially expensive to manufacture. Because it was magnetic, it was prone to interference. In the case of the core memory, adjustments with regard to the sampling level, the drive currents and the storage time were necessary. Time consuming applications were required to diagnose hardware problems in the core memory.