Computer memory

Computer memory is a vital component that stores data and programs for immediate use.

It holds the specific instructions a computer fetches and the data those instructions process. Scientists and engineers often call this main memory or primary storage. It is frequently referred to as RAM, which stands for random-access memory. This type of memory operates at a very high speed. This speed is much greater than mass storage, which is slower but cheaper per bit.

Because it is so fast, memory serves as a cache and a write buffer. This helps improve both reading and writing performance for the system.

Modern computer memory is implemented as semiconductor memory. This technology uses memory cells built from MOS transistors on an integrated circuit. Most of these cells store a single bit, represented as a 0 or a 1. These cells are grouped into words of fixed lengths. Common word lengths include 1, 2, 4, 8, 16, 32, 64, or 128 bits. Each word can be accessed by a binary address of N bits. This mathematical structure allows a system to store 2^N words in the memory.

This organized system allows the computer to find and use information almost instantly.

Semiconductor memory is divided into two main categories: volatile and non-volatile. Volatile memory requires constant power to maintain its stored information. If the power is lost, the data vanishes. Two main types of volatile memory exist: static random-access memory (SRAM) and dynamic random-access memory (DRAM). SRAM is used mainly for CPU cache and can use six transistors per bit. DRAM is used for primary storage and uses only one transistor and one capacitor per bit. This makes DRAM much denser and cheaper per bit. Non-volatile memory, such as flash memory or ROM, retains data even without power.

In the early 1940s, memory capacity was extremely limited. The ENIAC, the first electronic programmable digital computer, used thousands of vacuum tubes.

These tubes could only store about 20 numbers of ten decimal digits. J. Presper Eckert later developed acoustic delay-line memory. This used mercury in glass tubes to store bits as sound waves. In 1946, Fred Williams invented the Williams tube using electron beams.

This was the first random-access memory, though it was sensitive to environmental disturbances.

Significant progress occurred with the development of magnetic-core memory. Frederick W. Viehe and An Wang developed it in the late 1940s. Jay Forrester and Jan A. Rajchman improved it in the early 1950s. It was commercialized with the Whirlwind I computer in 1953. Magnetic-core memory was a major breakthrough because it was non-volatile. It allowed for memory recall even after a power loss. This technology remained the dominant form of memory until MOS semiconductor memory arrived in the 1960s.

The invention of the MOSFET enabled the practical use of MOS transistors for storage. John Schmidt developed MOS memory at Fairchild Semiconductor in 1964. This technology was cheaper and used less power than magnetic core. In 1966, Robert H. Dennard at IBM developed a single-transistor DRAM cell. He discovered that storing a charge on a MOS capacitor could represent a bit. This led to the first commercial DRAM IC chip, the Intel 1103, in October 1970.

This advancement paved the way for the high-density memory used in modern PCs.

Non-volatile memory also saw many specialized inventions. Wen Tsing Chow invented programmable read-only memory (PROM) in 1956. In 1971, Dov Frohman at Intel invented EPROM, which is erasable PROM. Later, EEPROM was developed in 1972 to be electrically erasable. Fujio Masuoka at Toshiba invented flash memory in the early 1980s. He presented NOR flash in 1984 and NAND flash in 1987. These developments allow us to store massive amounts of data permanently in very small devices. Today, these technologies connect to broader systems like mass storage and virtual memory management.