In an SSD, the information stored is written to semiconductor cells. These cells retain their status even when there is no power supply – the principle of flash memory. A single memory cell can only have two states: charged or uncharged. This method is called Single Level Cell (SLC) and is mainly used in very expensive, industrial SSDs. One cell corresponds to one bit which illustrates how many such cells are needed to realize one gigabyte (1 GB), for example 109 = 1 billion memory cells (exact value: 230 = 1,073,741,824). A single letter in ASCII coding alone consumes 8 bits. That might give you an idea of how much memory is needed for a text document or for images.
However, it is also possible to use different voltage sizes in one cell so that more than 1 bit can be stored per memory cell. This type of storage is called Multi Level Cell (MLC) and usually allows 2 bits per cell. So, more data can be accommodated in the space and it works out more cost-effective. Its disadvantage is the lower number of write cycles. Another compression step is called Triple Level Cell (TLC), which further reduces manufacturing costs.
Semiconductors have a limited lifespan. To counteract this, an SSD has internal monitoring that detects worn memory cells. This Bad Block Management marks cell blocks with declining memory cells as faulty and replaces them with cells from a reserve stock. Depending on the SSD design, this comprises two to seven percent of the total storage capacity and considerably extends the usage life of an SSD.
Hybrid hard disks (HHD), however, shouldn’t be forgotten. This is a combination of an HDD hard drive and SSD. The fast flash memory of the SSD can increase the overall speed of such a hybrid compared to normal HDDs but doesn’t come close to that of single SSDs.