Flash memory
From ArticleWorld
Flash memory is a type of an Electrically-Erasable Programmable Read-Only Memory (EEPROM) that facilitates the erasure or rewriting of several memory locations in a single programming operation. This memory device can be classed under the category of ‘Non-Volatile Random Access Memory’ (NV-RAM), since data can be written several times over by a user or programmer, as in RAM and at the same time avoiding data loss when power supply is disconnected, as in ROM. It has earned the name ‘flash’ because data stored can be erased and rewritten in a matter of seconds. Flash memory is mainly used in USB flash drives, MP3 players, cellular phones and digital cameras.
How flash memory works
Memory is stored in arrays of floating gate transistors. Information in the form of single or multiple bits can be stored in these transistors, which are often referred to as cells. The capacity of flash memories may vary from a few kilobits to several gigabytes depending upon the application.
In flash memories using the NOR gate, two gates, namely, the control gate (CG) and the floating gate (FG) are present. The floating gate is installed in between the control gate and the substrate of the cell. Electrons arriving at the FG are held on there as a layer of insulating oxide prevents their escape. In a process termed as hot-electron injection electrons at the FG are sucked into their appropriate position by a strong electric field on the CG. This is how information is stored in NOR-based flash memories.
Every cell in the memory has a threshold voltage. Whenever it is required to read from the memory chip, electric current will be caused to flow depending upon this voltage. Ultimately, it is the concentration of electrons at the FG which decides the value of this voltage. The flow of current or the lack of it is detected and finally translated into binary values of 1’s and 0’s. In the case of multi-level cells the magnitude of current flowing is used to interpret stored data.
If case data has to be erased for ‘flashing’ the memory, a large potential difference is placed between the source and the CG. This results in electrons being withdrawn from the FG. All cells present in an entire block can be erased simultaneously. Programming of such memory chips can be done either byte by byte or word by word.
Apart from NOR memories, NAND-based memories are becoming common nowadays. In NAND memories, a process called tunnel injection is used to pass electrons through a thin layer of an electrical insulator. The process of erasing involves a technique called tunnel release. Programming and erasing these chips is easier.
Development of flash memory
Flash memory was invented by Toshiba in 1984, with Dr. Fujio Masuoka the pioneer behind it. Just a few years later, Intel began to produce the device to be sold commercially. In the beginning, NOR-based flash chips were popular, with their erase-rewrite cycles numbering up to several hundred thousand. Such chips were suitable for computer BIOS memories.
Later devices saw the increasing use of the NAND flash memory because of lower costs involved, faster erase-read cycles, greater density and improved endurance. The interface of such chips being sequential, they are considered good for large-capacity data storage. Models like those of Reduced Size Multi Media Card (RSMMC) and the TransFlash have greatly reduced format sizes.
It has been suggested that flash memory of capacities of several gigabytes can be made to replace hard disk drives. This seems a little far-fetched at the moment because of the fact that flash memory involves longer time while erasing data; an entire block of memory has to be done away with. The programming cycles are limited to one million. Also, flash drives are prohibitively expensive.