Flash Storage Configurations: NOR and NAND

(Source: afxhome/stock.adobe.com)

Plugin solid-state memory devices are based on flash memory technology invented by Toshiba's Fujio Masuoka (1943-) in 1980. A flash memory device consists internally of large arrays of floating-gate metal-oxide semiconductor field-effect transistors (MOSFETs). It comes in two (2) configurations:

• NOR (Not OR)
• NAND (Not AND)

NOR-based flash has fast-read access times and high endurance. Because it allows random access to any memory location, it is used mostly for reliable code storage in ROM-replacement BIOS or firmware applications. NOR devices are available up to 2GB, with either a serial or parallel interface.

NAND flash requires less chip area per cell, allowing greater storage density and lower cost per bit than NOR flash. NAND devices are available up to 2Tb. The disadvantage of NAND flash is that individual cells can’t be written or erased; these operations must be done in blocks of cells, making NAND more suitable for data-heavy applications. The write operation slowly degrades the NAND cell; most NAND drives wear out after a few thousand write cycles, whereas read operations can extend to millions of cycles. This limitation still gives a typical lifetime of several years under normal operation.

The process geometries for NAND flash are as small as 25nm. There are three (3) different types of technologies: Single-Level Cell (SLC), Multi-Level Cell (MLC), and Triple-Level Cell (TLC), which can store one, two, or three bits per memory cell, respectively. As the bits per cell increase, the cost decreases, but the performance and endurance degrade. Higher-performance products, therefore, tend to use SLC and MLC devices, while consumer-grade devices are moving to TLC technology.