A flash cell stores one bit by trapping electrons on a floating gate. Take apart the transistor stack — control gate, floating gate, tunnel oxide, channel — and see how a tiny charge becomes a 0 or a 1.
A NAND flash cell is the tiny transistor that stores one bit of data in an SSD, a USB stick or a memory card. It is a normal transistor with one extra trick: a conducting layer in the middle, called the floating gate, that can hold electric charge.
Because that charge does not need any power to stay there, flash is non-volatile — it keeps its contents when the power is removed. Billions of these cells, wired into rows and columns, form the storage inside an SSD.
From the outside, a flash memory chip is a black plastic package with no visible detail. Everything that matters is a horizontal sandwich of thin layers, so a cutaway is the only way to see how a voltage writes, reads and erases a bit.
Writing starts with a voltage on the word line, the row of metal that connects the control gates of a whole line of cells. This lifts the control gate and creates a strong electric field across the stack below it.
The field is strong enough to push electrons straight through the very thin tunnel oxide that separates the channel from the floating gate. In normal use that oxide blocks current, but at writing voltages it lets electrons pass.
The floating gate is a conductor surrounded on all sides by insulators — the tunnel oxide below and the blocking oxide above. Once electrons arrive, they have nowhere to go, so they stay put, held in place by the insulating layers.
The electrons on the floating gate change the voltage needed to make the transistor conduct. A cell holding charge reads as one value and an empty cell reads as another — that difference is the stored 0 or 1.
To read the cell, a small voltage is applied across source and drain. If the floating gate holds charge, the channel stays closed and little current flows; if it is empty, the channel conducts. The cell is never destroyed by reading.
Erasing applies a reverse field that drives the trapped electrons off the floating gate and back towards the channel or substrate. Because many cells share the same region, erasing is done in large blocks rather than one cell at a time.
No. The charge sits on a floating gate that is completely surrounded by insulators, so it drains away only very slowly over years. That non-volatile behaviour is why SSDs, USB sticks and memory cards keep their contents without power.
A conducting layer in the middle of the transistor that is not connected to anything. Electrons placed on it are trapped by the insulating oxide above and below, so the charge stays put and can be detected later.
Erasing means removing charge from many cells at once through the shared substrate or word line, so it is done block by block. Writing, by contrast, can target single pages, which is why flash is described as written in pages but erased in blocks.
Each program and erase cycle pushes electrons through the tunnel oxide and slowly damages it. After enough cycles the oxide leaks, so the cell no longer holds its charge reliably — this is why SSDs have a limited write endurance.
A simple single-level cell stores one bit: charged or not. Storing more bits per cell means placing several distinct charge levels on the floating gate, which is how MLC and TLC flash get more capacity at the cost of speed and endurance.