Most people understand — at least vaguely — that hard drives can fail. The clicking noise, the grinding, the drive that spins up but never mounts: these are familiar signs of something going wrong mechanically. SSDs are different. They fail quietly, often without warning, and the physical process that causes them to degrade is one that begins the moment you start using them. Understanding how NAND flash memory wears out, what TBW ratings actually mean, and why SSDs fail so differently from hard drives is increasingly important as SSDs become the primary storage medium in laptops, desktops, and external drives across South Africa and globally.

How NAND Flash Wears Out: P/E Cycles Explained

Every SSD stores data in NAND flash memory cells. Unlike a hard drive, where data is written to a magnetic platter that can be overwritten indefinitely, NAND cells have a finite number of program/erase (P/E) cycles — a complete write-and-erase operation degrades the cell’s insulating layer slightly each time. Enough cycles and the cell can no longer hold a reliable charge, meaning it can no longer store data accurately.

The number of P/E cycles a cell can tolerate depends on how many bits it stores. SLC (single-level cell) NAND stores one bit per cell and tolerates around 50,000 to 100,000 P/E cycles — extremely durable, but expensive per gigabyte and rare in consumer products. MLC (two bits per cell) tolerates around 3,000 to 10,000 cycles. TLC (three bits per cell), which is now the standard in most consumer SSDs, manages around 1,000 to 3,000 cycles. QLC (four bits per cell) — increasingly used in high-capacity consumer drives — drops to around 100 to 1,000 cycles per cell.

TBW: The Lifespan Metric Most Buyers Ignore

TBW — terabytes written — is the manufacturer’s rated lifespan for an SSD expressed as a total write volume. A drive rated at 300 TBW can have 300 terabytes of data written to it before the NAND cells are expected to approach end-of-life. A drive rated at 600 TBW can handle twice as much. Most buyers compare capacity and price per gigabyte. Few check the TBW rating, which is actually the more meaningful number for understanding how long the drive will last under real-world use.

For a typical home user writing 20–40GB per day — which covers documents, downloads, browser cache, and occasional file transfers — a 300 TBW drive would last 20 to 40 years at that rate. In practice this means most consumers will replace their drive for other reasons long before reaching the TBW limit. However, users doing video editing, software development, database work, or regular large-file transfers can consume TBW much faster, and in those cases the rating matters considerably.

The picture changes with QLC drives. A 2TB QLC SSD might carry a 360 TBW rating — seemingly reasonable — but at 200GB of writes per day, that is just under five years of headroom. For a workstation drive used in any write-intensive capacity, that is a meaningful constraint.

The QLC Shift — More Capacity, Less Endurance

The storage industry has been moving steadily toward QLC NAND as manufacturers pursue higher capacities at lower cost per gigabyte. This is good news for consumers who want large, affordable SSDs. It is a more nuanced story for reliability. QLC drives perform well under typical read-heavy workloads and light-to-moderate writing — the kind of use most home and office computers actually see. But they are more sensitive to write-intensive use, higher temperatures, and extended periods of low charge (a phenomenon called data retention degradation, where stored charge leaks from cells over time if the drive is left unpowered for months or years).

The practical guidance for anyone storing important data: check the TBW rating before buying, prefer TLC over QLC for any drive that will see sustained writes, and never rely on a single drive — SSD or HDD — as the only copy of data you cannot afford to lose.

Why SSD Failure Is Silent — and Why That Matters

Hard drives often give warning before they fail. SMART data shows reallocated sectors climbing. You hear a click or a soft grinding. The drive slows noticeably. These are signals that something is wrong and that you should back up immediately. SSDs frequently give no such warning. A drive that is working normally one day may be completely unresponsive the next — not because something catastrophically broke, but because the controller has detected the NAND has reached a threshold and has locked the drive to prevent further writes or reads.

Some enterprise SSDs handle this gracefully — they enter a read-only mode at end-of-life, allowing data to be copied off before the drive is retired. Consumer SSDs are less consistent about this. Many simply stop responding entirely when the controller detects too many uncorrectable errors, giving the user no time to react.

The other common failure mode is controller death from power events — a sudden power cut, voltage spike, or unclean shutdown during a write operation can corrupt the controller’s mapping table, making all the data on the drive inaccessible even though the NAND cells themselves are physically intact. This is one of the most recoverable SSD failure types in a specialist lab, but it is also one of the most misunderstood — many users (and many IT shops) assume an unresponsive SSD means all data is gone.

SSD Data Recovery: What Is and Is Not Possible

SSD recovery is fundamentally different from hard drive recovery. There are no platters to image, no heads to replace. Instead, recovery involves working with the controller, the firmware, and in more complex cases the NAND chips themselves. The most common recoverable SSD scenarios are firmware and controller failures — where the drive has lost its mapping table but the data on the NAND is physically intact — and logical failures such as accidental deletion, corruption, or formatted partitions.

Genuinely worn-out NAND — a drive that has exhausted its P/E cycles — is the hardest scenario. Once cells can no longer hold a reliable charge, recovering data from them requires reading at a cell level and correcting errors across thousands of damaged cells, which is extremely difficult and not always successful. This is why the most important piece of advice for SSD owners is straightforward: do not wait until the drive fails. Monitor TBW using the manufacturer’s SSD management software, watch for SMART warnings, and maintain a backup.

If your SSD has already failed — whether after a power event, a sudden unmount, or gradual degradation — SouthBit provides SSD data recovery for all types including SATA, NVMe, and M.2 drives. Clients across South Africa send drives to our Cape Town lab by overnight courier, including from Bloemfontein, Gqeberha (Port Elizabeth), Johannesburg, and Durban. The assessment is free and there is no charge if recovery is not possible.

Sources:
ATP Electronics — SSD TBW, DWPD and Endurance Explained  · 
Newegg Insider — SSD Endurance & Reliability: What Buyers Should Know  · 
OSCOO — How Long Do Solid State Drives Actually Last?  · 
Backblaze Drive Stats for 2025