Data Recovery Case File · SSD · An Honest Limit

0 bytes of 0 bytes: an 870 EVO's sudden lockout, an identical twin on the desk, and the swap that must not be tried

The failure had the abruptness SSDs specialise in: the drive appeared, transferred files normally, was ejected properly — and never showed up again, reporting on direct SATA connection a total capacity of 0 bytes. What makes this enquiry worth its own page is what sat next to the dead drive: an identical 870 EVO, same capacity, same firmware version, working perfectly. Two matching patients, one alive — and the most dangerous temptation in DIY SSD recovery.

The enquiry

“I have two identical 1TB Samsung 870 EVO SSDs with the same firmware version; one has failed. It appeared and I transferred data successfully, then disconnected it. Now it doesn't show up at all — connected via SATA it displays 0B written and 0B total capacity in diskpart, and doesn't appear in Partition Manager. I have not performed any other operations on the drive.”

What "0 bytes" means on this drive

A drive announcing zero total capacity hasn't been erased — it has stopped being able to describe itself. The controller wakes, fails to bring its internal mapping of the flash online, and retreats into a failsafe presentation: alive enough to answer, honest enough to claim nothing. On this Samsung family it's a documented way for the drive to die, arriving — as here — without a single warning symptom, mid-life, after a perfectly normal session. The customer's data is not destroyed; it's sealed behind machinery that no longer runs, on NAND the controller additionally encrypts as a matter of design. Which frames the honest question: can anyone restart that machinery? Our assessment, given free and in writing, was the one honesty allows: Samsung's in-house controllers — SATA and NVMe alike — are a closed world — no third-party technological access, no public firmware toolset — and reading the chips directly around the controller yields ciphertext only that controller's machinery could interpret. We won't sell hope against that wall, and we'd urge scepticism toward anyone who does without explaining precisely how.

The twin drive — and why it must stay in its lane

Now the temptation his identical spare creates, addressed head-on because forums recommend it weekly: swapping boards or components between a matching pair. On hard drives of a certain age that logic once held; on an SSD it is precisely backwards. There is no "board" separate from the storage — controller, firmware state and flash are one interdependent system, and the map of where every scrap of data physically lives exists nowhere but in the failed drive's own private records, encrypted to its own keys. Transplants between twins don't revive the patient; they risk the donor and can overwrite the very structures any future recovery would need. His restraint — "I have not performed any other operations" — was, as with the best customers in these files, the most valuable technical contribution available. The drive left as it arrived: intact, unpowered, its options preserved, with our written advice to keep it exactly so, since sealed is not the same as gone and tooling landscapes do shift — a possibility we note without promising.

Outcome

No recovery, no fee, no varnish — and the working twin now backed up on a schedule, because the sharpest lesson in the room was standing next to the failed drive wearing the same label. Identical drives, identical firmware, identical duty: one dead without ceremony. SSDs don't fail like their siblings warn you; they fail like this page. The second copy is the entire defence.

Our case files are drawn from genuine enquiries received by our laboratory over the past ten years, anonymised to protect client confidentiality. Each one describes the diagnostic and recovery procedure our engineers apply to that fault, using the equipment listed.