Flash Memory - what is it? how does it work ?

Flash memory relies on storing bits as electron charges on the floating gates of MOSFETs. These are dual gate devices and the other gate is used to control storing or draining the charge. Basically you have lots of very small capacitors that are either storing a small charge, or not, along with a means of either, reading the charge state of each capacitor without affecting its charged state, or changing the charge state when required.

A power supply is required to read, erase or program Flash memory, but it will 'remember' its contents when power is disconnected. It is non-volatile memory (well relatively). Flash memory chips are packaged in a number of different physical formats, for example

- Compact Flash cards and USB memory sticks.

Flash memory is so called because the entire sections of the microchip are erased at once or (flashed). The erasure is caused by Fowler-Nordheim tunnelling. Although you can program or read Flash one 'word' at a time, erasures have to be a block at a time. Erase/program cycles are limited to around 10 to the 6 times. Although capacities are now up to 32GB, most commercial use is around the 128MB to 1GB levels. Flash is a more idiot friendly name for what engineers used to call EEPROM or Electrically Erasable Programmable Read Only Memory.

For more in depth info see - http://en.wikipedia.org/wiki/Flash_memory