Data Recovery Case File · Second Fixes & Trade Handoffs · Ask What Was Extracted
The Chip Came Off, and the Question Is What Came With It
His enquiry follows the most invasive procedure there is for flash storage. A stick where "a specialist said the issue was a failed memory chip and attempted a chip-off procedure. Unfortunately they were unable to recover any data in this way due to 'data damage' and have said there are no other options." The work is needed to finish a degree. "Data damage" covers two situations that are not remotely alike — and which one applies decides whether there is anything left to attempt.
| Media | USB flash drive following chip-off extraction by a previous provider — raw memory read; reconstruction unsuccessful; device no longer intact |
| Reported situation | Data lost from a flash drive · assessed by a specialist provider · memory chip failure diagnosed · chip-off extraction performed · no data recovered · damage to the data cited as the reason · no further options offered · academic work required |
| Fault class | Post-extraction reconstruction failure — degraded memory and unresolved translation layer producing identical descriptions; raw extraction retention determinative |
| Equipment used | Retention of the previous provider's raw extraction established before any other step · descrambling, error-correction and translation-layer reconstruction attempted independently against that extraction · physical memory re-read where the chip was retained · position stated honestly before any charge |
The decode: what chip-off produces, and what has to happen next
What the procedure actually does: removes the memory chip from the board and reads it directly on dedicated equipment, bypassing the controller entirely. It is the last resort and it is irreversible — the stick no longer exists as a device afterwards, and the chip is the only remaining object.
What comes out of it, and this is the part people do not expect: not files. Not a filesystem. A raw dump of the memory in the order the chip stores it, which bears no resemblance to the order a computer would have read it. Flash controllers scramble data to spread wear, interleave it across planes, and apply error correction — all according to algorithms specific to that controller.
So the extraction is the beginning, not the end. The dump must be descrambled, error-corrected, de-interleaved and reassembled using a reconstruction of how that particular controller organised things. That step is where most chip-off work succeeds or fails, and it is entirely separate from whether the chip read cleanly.
Which is why "data damage" is ambiguous. It might mean the memory itself is degraded — cells whose stored charge has drifted beyond what error correction can reconstruct, which is genuine and final. Or it might mean the extraction was clean and the reassembly did not resolve, which is a different problem entirely and one that another attempt can address.
So the question that decides everything: did they keep the raw extraction? If they retained the dump, a second attempt works from it directly — no further extraction needed, no risk to anything, and the whole question becomes whether the reconstruction can be resolved. If they did not, and they retained the chip, it can be read again. If neither was kept, the position is much harder.
Why asking is entirely reasonable: providers routinely retain extractions for a period, and the request is a normal one. It costs him nothing and it may be the whole case.
Being honest about the odds: a specialist who performed chip-off did serious work, and their finding may be exactly right. Where the memory is genuinely degraded, a second attempt confirms it and should not be charged for. The reason to ask is that the alternative reading is common enough to be worth eliminating.
On the bench
Retention of the previous provider's raw extraction was established before any other step — chip-off producing a raw dump requiring descrambling, error-correction, de-interleaving and translation-layer reconstruction, which is where such work most often fails and which is separate from whether the memory read cleanly. Reconstruction was attempted independently against that extraction, and the physical memory re-read where the chip was retained. The position was stated honestly before any charge.
The outcome
The previous extraction sought before anything else, reconstruction attempted independently, and the position stated before any charge. Free assessment, one fixed written figure including VAT, 50% of parts and labour upfront with the balance only on successful recovery — and no charge where nothing further is possible. The decode: chip-off produces a raw dump, not files. It has to be descrambled, error-corrected and reassembled according to how that controller organised things, and that step is where most such work fails. So ask whether they kept the extraction.
After a chip-off that recovered nothing
Ask whether they kept the raw extraction — that single question may be the whole case, and requesting it is entirely normal. Chip-off removes the memory and reads it directly, but what comes out isn't files: it's a raw dump in the order the chip stores things, scrambled to spread wear, interleaved across planes and error-corrected according to algorithms specific to that controller. It has to be descrambled and reassembled afterwards, and that step is where most chip-off work actually fails. So "data damage" might mean degraded memory, which is final, or an unresolved reconstruction, which isn't. If they kept the dump, a second attempt works from it with no risk at all.
Ask if they kept the extraction — then call Oxford Data Recovery on 01865 593000; reconstruction attempted independently against the raw dump, memory re-read where the chip was retained, honest position before any charge.
Request a quote online →
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.