The modern Solid State Drive (SSD) operates on entirely different physical principles than conventional magnetic storage, relying on NAND Flash memory to provide fast, non-volatile persistence. The fundamental storage unit is the NAND cell, but data is read and written in fixed-size pages (typically 4KB to 16KB), and crucially, can only be erased in much larger units called blocks. This read/write asymmetry creates inherent management challenges for the drive. Before a single page within a block can be written or rewritten, the entire block must first be copied to a buffer, erased (an electrically intensive process), and then written back with the modified page.

This process is governed by the Flash Translation Layer (FTL), which is an embedded software and hardware component within the SSD controller. The FTL is responsible for translating the logical block addresses (LBA) used by the host operating system into the physical block and page addresses within the NAND chips. This translation is not static; the FTL constantly employs sophisticated algorithms for wear leveling to distribute write operations evenly across all blocks. Since NAND cells have a finite number of erase cycles, wear leveling is essential for maximizing the drive’s lifespan. Furthermore, the FTL must manage garbage collection, a background process that identifies blocks containing mostly stale or invalid pages, consolidates the few remaining valid pages into a new block, and then erases the old block. This complex housekeeping leads directly to write amplification, where the amount of physical data written to the flash media exceeds the logical amount of data written by the host, representing a continuous management overhead necessary to preserve both performance and longevity.