Call us — 01483 901310
Mon–Fri · 9am–5:30pm · No fix, no fee
Start a free diagnostic →
// case file · SSD · Samsung EVO M.2 NVMe · thermal failure

A Samsung NVMe that overheated its own controller.

An architecture practice’s seven years of CAD work, the lot on one M.2 stick that had run too hot for months — until it threw off a burning smell and just stopped being there.

← All case files · £300 + VAT, flat

Device

Samsung EVO M.2 NVMe SSD

Failure

Thermal — controller burnout, board damage

Data

7 years of CAD drawings, project files

Outcome

Full CAD archive returned

// the brief

What turned up, and what was on the line.

The workstation drive from an architecture firm. It had been sitting far too hot even under everyday load — the one warning everybody overlooks, since a hot drive doesn’t read as a dying one. Then came a burning smell, and the SSD vanished from the BIOS completely. Not corrupt — missing. The slot just showed empty.

On board: seven years of drawings, CAD files and project data, not a scrap of it backed up. CAD is a particularly cruel thing to lose, because a 3D assembly has no in-between state — it’s either intact or it’s rubbish.

// on the bench

What the diagnosis showed.

01

The controller had burnt out.

these drives force a great deal of data through a minuscule chip, and in a workstation whose M.2 slot had no cooling whatever, the heat simply mounted. Kept under that thermal strain long enough, the controller die and the solder beneath it break down.

02

The board carried thermal damage

around the power-delivery circuitry.

03

The NAND had survived.

the finding that counted, and the one we nearly always meet — the memory chips are perfectly sound yet impossible to read, because the record of where each block physically sat was held by the controller, not on the chips.

// the recovery

How we did it.

With the controller dead, we bypassed it and went straight for the memory. The NAND chips were read off the board on a flash programmer for a raw dump, and then the real work began: reversing the precise scheme the controller had imposed — its error-correction and the way it scrambled and spread data over the chips — to turn that dump back into a coherent image. From there the file system was rebuilt and the data written out to a fresh drive. This is exactly the flash-level work our programmers and the PC3000 exist for.

// outcome

What we recovered.

With the NAND read and the translation layer reassembled, the drawings were handed back — blueprints, 3D models, project files and the whole historical design archive — on a fresh drive.

A burnt controller sitting over sound NAND makes that the outcome you’d expect, not a fluke. What truly makes flash recovery a gamble isn’t how dire the damage looks — it’s the controller family. Some are well documented and their translation layers rebuild cleanly; others aren’t, and a handful of recent designs fuse controller and NAND so tightly that there’s nothing to lift at all. That’s something the free diagnosis settles, not a phone call.

// the transferable bit

What to take from this.

The warning sign here is thermal throttling, and it’s the only one an SSD offers. When a drive begins running hot and slowing down, that’s it telling you — in the one way it’s able — that it’s under strain. No click to catch, no grinding, no graceful fade. One day it works; the next it’s simply not there.

So if a machine has been overheating and stammering under load, don’t save the backup for the weekend — do it the same day. And wherever an M.2 drive sits exposed in a workstation or a slim laptop with nothing to cool it, spend the handful of pounds a heatsink costs and fit one.

// read next

Related.

// your turn

Lost something that matters? Free diagnosis, a fixed price, and no fix, no fee.

Bring the drive to our Guildford Business Park reception, or post it over — it costs nothing to learn what went wrong. You’ll have a written figure from the fixed bands before any work begins.