Flash in Devices

A series on the flash storage built into devices rather than sold as a drive. Part 1 covers UFS, the flash storage inside a phone: its layers, how it differs from eMMC and NVMe, the link gears, each version up to UFS 5.0, command queueing, how a read reaches the flash, and UFS cards and cars. Part 2 covers eMMC, the managed NAND on a parallel bus that phones used before UFS: its eleven wires, speed modes up to HS400, partitions, command queueing, real parts and why makers still sell it. Part 3 covers SD and microSD cards: the capacity families, bus modes from Default Speed to UHS-III and SD Express, speed classes, what is inside a card, endurance cards and fake-capacity cards. Part 4 covers automotive and industrial flash: temperature grades and AEC qualification, retention, endurance under constant logging, health and power-loss protection, the Tesla eMMC recall, and the move from eMMC to UFS and BGA SSDs in cars. Part 5 covers flash in phones and tablets: PoP, ePoP and uMCP packages, storage tiers that come with fixed RAM, and what Android’s memory manager, F2FS and file-based encryption do with the flash. Part 6 covers flash in small devices: SPI NOR and SPI NAND in routers, cameras, smartwatches and microcontroller boards, how a chip is split into partitions, and how bad blocks, wear and power loss are handled without a controller. How a flash cell stores a bit is in Flash Foundations; these guides start at the package and the interface.