With SSDs people talk about how many writes they last and about abbreviations like TLC or QLC. It relates to memory density: the more data we cram into the same space, the cheaper the drive, but usually the less durable too. Let us explain how it works for SSDs and for classic drives.

NAND cells: as many bits, such a lifespan

An SSD stores data in NAND memory cells. The key is how many bits are stored in one cell. More bits mean higher density and a lower price per gigabyte, but fewer possible rewrites, that is a shorter lifespan.

TypeBits per cellDensity and priceLifespan (P/E cycles)Use
SLC1lowest, most expensive~50,000-100,000industry, servers
MLC2lower~3,000-10,000pro, older drives
TLC3high~1,000-3,000ordinary quality SSDs
QLC4highest, cheap~100-1,000cheap, high-capacity SSDs

The values are approximate. PLC (5 bits) is also arriving, even denser and even less durable. A P/E cycle means one write and erase of a cell.

3D NAND: layering solves capacity, not lifespan

Modern SSDs store cells in vertical layers (3D NAND), today commonly 96, 128 and 232 or more layers. This layering increases mainly capacity, without the cells having to shrink extremely. That is precisely what helped keep the lifespan at a reasonable level.

So two things must be told apart: the number of layers mainly solves capacity, the number of bits per cell (SLC to QLC) mainly solves lifespan.

How an SSD’s lifespan is stated and protected

  • TBW (Terabytes Written) says how many terabytes can be written to the drive during the warranty.
  • DWPD states how many times a day the whole capacity can be rewritten.
  • The drive helps itself with techniques: wear leveling (even wear of all cells), reserve capacity (overprovisioning) and an SLC cache for fast writes. There is more on the inside of an SSD in the article DRAM cache in an SSD.

An important reassurance: an ordinary user does not really “write through” a modern TLC or QLC drive in years. The write limit only matters in heavy deployments.

And what about an HDD: layers are platters

In a classic drive (HDD) the “layers” are actually platters, and density is given by the areal density of the recording. Important: an HDD has no write count limit like flash. It wears out mechanically, that is by hours of operation and head movement.

Higher capacity comes from technologies like helium drives, more platters and SMR (overlapping tracks). SMR, however, slows down random and repeated writing. So it is not a “maximum number of writes”, but a performance and rewrite issue. The differences between drive tiers and CMR versus SMR are covered in the article WD and Seagate drive tiers.

What to take from this when choosing

  • For a lot of writes (servers, caching, video editing) choose more durable SSDs, TLC rather than QLC, or enterprise, and watch the TBW.
  • For ordinary use and archiving TLC or QLC is perfectly fine.
  • Watch the drive’s health (SMART), more in the article SMART warnings on a drive.

Conclusion

For SSDs the rule is that the more bits per cell, the higher the density and the lower the price, but the fewer writes and the shorter the lifespan (SLC, MLC, TLC, QLC). 3D NAND layering mainly solves capacity. An HDD has no write limit like flash, it wears out mechanically, and higher density via SMR rather harms its performance. When choosing, go by how much you will write and check the TBW.

Choosing a drive for a computer, server or NAS and want to pick the right one? Get in touch, we will advise according to your use and budget.

This article is part of our Hardware and components overview.