User:Abujaki

Essay:

How does it work?

The data is stored as a certain voltage either above a preset x level, or below a preset y level, both determined by the manufacturer. A high voltage is typically read as a 1, and a low voltage is typically a 0. The voltage in question is preserved in a 'Floating gate' transistor, where an additional gate is sammiched between two insulators. This ensures there is no voltage leak (Provides the nonvolitility and prevents damage to the stored data during the read step.)

NOR and NAND

The two main types of flash memory are NOR and NAND. Both of them consist of a grid of memory Floating Gate transistors.

NOR

The chips are arranged in a logical grid pattern in paralell, using Word lines and bit lines to access the transistors to read/write to. Word = Y lines Byte = X lines. • Reads faster • Should be careful not to over-erase the transistors (Depletion mode) • Fast random read/write • Additionally, NOR Guarantees 100% good bits, so error-checking and correction doesn't need to be programmed into the system.

NAND

* 8 or 16 transistors in series, between two select gates. * The string gate ensures selectivity * The Ground gate keeps current from passing through the cell while the memory is programming * Depletion mode is okay, but the * Physically smaller cell size (85% smaller than NOR) [20] * Because the bytes are in serial, a word can be written to memory all at once instead of bit-by-bit like in NOR * Primarily preferred for data storage and removable media * Offers redundant memory and error-correcting to remove bad bits. * Comparitively slower read times because the bytes are linked in serial. * More complex than NOR * Slightly more costly than NOR

Strengths

* Cheap * High-density storage (Lots of data per square inch) * Long data retention * Fast read/write times * Erasable * Flash memory is nonvolatile. This means even after the power is removed from the device, the memory device will retain the information programmed into it. Typical manufacturer specifications require nonvolatile memory to retain data for at least 10 years before the first errenous readout of data.

Weaknesses

* Limited read/erases. (See endurance) * Not Byte-alterable. In order to flip a byte, you would need to erase and reprogram an entire memory block

Why use flash?

* Flash memory, because of its data retention capabilities coupled with quick read/write cycles is a staple in today's portable device technology. * NOR Flash memory quickly replaced EPROMs ( Erasable Programmable Read-Only Memory) in development. While both are nonvolatile and erasable, Flash also offers cost-effectiveness, both to manufacture and replace, was high-density, and offered fast reads.

Data Retention

Data retention refers to the period of time between storing the data, and the first errenous readout of the data. Typical memory specifications require flash memory to be able to retain data for 10 years.

Endurance

Endurance refers to the number of write/erase cycles the memory can endure before the technology is worn out, due to stress from reads and erases. Over time, the retention capability of the memory will degrade, and A general industry requirement is 100,000 erase/write cycles before failure.