Digital storage is kind of pointless if we cannot know exactly which bit came from what logical location.
ie: it's not like analog audio tape, where we get some signal, but don't really know its exact position.
Digital storage is kind of pointless if we cannot know exactly which bit came from what logical location.
ie: it's not like analog audio tape, where we get some signal, but don't really know its exact position.
With all flash, a cell is the smallest addressable unit. Bits are stored in cells and depending on the flash type it can store anywhere from 1 to 5 bits per cell. So with SLC flash a cell is analogous to a bit, but for the multi-cell flash types a cell might mean 2, 3, 4 or 5 bits.
Flash cells are organized into strings, pages, blocks, planes, and then the flash die itself. Reading and writing of NAND flash occurs at the page level (4-16 KiB per page) but erasing occurs at the block level. So in the case of NAND, the controller cannot read/write a single bit of data.
So while you can specify that only 1 bit be changed in flash storage, there's an entire abstraction layer between you and the flash chip that hides the fact that a lot of data gets moved around to support a single bit being changed.