In the 1980s, Fuijo Masuoka invented flash memory while working for Toshiba. According to the Japanese multinational, the name ‘flash’ was suggested by Masuoka’s colleague, Shoji Ariizumi, who likened the erasure process of memory to a flash on a camera.
Developed from EEPROM (electrically erasable programmable read-only memory), Masuoka and colleagues presented the invention at the IEEE 1984 International Electron Devices Meeting (IEDM) held in San Francisco. Following Toshiba’s introduction of the technology to market in 1984, chip giant Intel was one of the first companies to see the potential of flash technology and released the first commercial flash chip in 1988.
How it works
Although derived and developed from EEPROM, flash is memory which is non-volatile. This means memory is still stored even if there is no power to the system. Memory that ‘forgets’ when there is no power is called Random Access Memory. In contrast to EEPROM, flash erases whole blocks of data at a time, rather than per byte.
Non-volatile memory is also synonymous with solid-state storage – another moniker attributed to flash. Solid-state storage is computer storage which involves no moving mechanical parts like you would see in a hard drive- instead storing and retrieving digital information using only electronic circuits.
Data is stored using electricity in surface-mounted chips on a PCB, or Printed Circuit Board, with the lack of moving parts making such a storage system ideal for more rugged applications, as well as reducing power consumption.
Data is stored via an array of memory cells made from floating-gate transistors. These floating-gate transistors, or memory cells, are similar to a standard MOSFET, but have two gates instead of one. Single-level cell (SLC) devices store one bit of information per cell, whereas multi-level cell (MLC) devices and triple-cell devices, can store more than one bit of information per cell.
A flash memory cell
Types of flash
The two main types of flash memory take their names from the NAND and NOR logic gates. NOR is the more expensive, less durable and slower memory of the two, with NOR’s larger individual cell size the reason for the memory being more expensive per gigabyte. Often seen as a good replacement for EEPROM, NOR uses no shared components and can connect individual cells in parallel, which enables random access to data. NAND flash cells are much more compact, with its fewer bit line architecture making NAND better suited to serial access, as opposed to random data access.
NOR erases data at the block level and rewrites data at the byte level. NAND, on the other hand, rewrites data at the multiple-byte page level.
NOR flash used primarily in consumer and embedded devices for boot and read-only code-storage use. NAND is usually found in consumer devices and enterprise storage systems due to its lower cost, greater density and higher programming and erase speeds.
Benefits of Flash
Flash is the cheapest form of semiconductor memory. The technology delivers higher performance than HDDs and tape, and has lower power consumption when compared with dynamic random access memory (DRAM) and static RAM (SRAM).
NOR flash is fast with random reads, while NAND delivers fast serial reads and writes.
Disadvantages of Flash
Flash does not last forever. Although manufacturers have improved endurance and shelf life, flash has a finite number of program-erase cycles.