Raw NAND ships with thousands of bit errors per page and would be unusable without error correction. This is the story of how ECC evolved from single-bit Hamming through BCH to soft-decision LDPC, what "soft bits" physically are, and the UBER math that turns a noisy die into a reliability promise.
Ssd
-
ECC for Flash: Hamming to LDPC -
NAND Interfaces: ONFI, Toggle, and the Speed Problem The wire between the flash die and the controller has quietly become the thing that bounds SSD throughput. A tour of ONFI and Toggle Mode, the climb from asynchronous SDR to 3600 MT/s NV-DDR3, what read/write training actually does, and why signal integrity now dictates how many dies you can hang off a channel.
-
Power-Loss Protection When the rail collapses mid-write, an SSD has milliseconds to harden its volatile cache and its mapping tables before the lights go out. A tour of the hold-up energy problem, capacitor banks and 1/2 C V^2 sizing, what "data at rest" weasel-wording actually covers, the dangerous SLC-to-TLC folding window, and how to test PLP claims with a relay rig instead of trusting the datasheet.
-
The Flash Translation Layer The FTL is the firmware lie that lets your SSD pretend to be a disk. A tour of L2P mapping, garbage collection and the write-amplification math, why overprovisioning is the master knob, how wear leveling and TRIM fit, and why fresh-out-of-box benchmarks are marketing and steady state is the only truth.
-
Zoned Storage and FDP Conventional SSDs spend DRAM, spare capacity, and write endurance fighting a problem the host created: blind data placement. Zoned Namespaces, Flexible Data Placement, and their HDD ancestor SMR all attack that waste by letting the host tell the drive where data goes. Here is how each one works, what it costs in software, and who should actually care.
-
How Data Lives on Platters and Flash: The Physics of Storage Magnetic domains, GMR read heads, perpendicular recording, SMR track geometry, NAND floating gates, charge-trap 3D NAND, wear leveling, and why all of this matters when you pick drives for ZFS — the physics underneath every storage decision.
-
fio Benchmarking Cookbook: Stop Using dd, Start Measuring What Matters Why `dd` tells you almost nothing about storage, and how to measure what matters with fio — building realistic jobs for IOPS, throughput, and latency across queue depths so your numbers reflect real workloads rather than cache.