Will it work? BiFeO3

It is worse than that - it is garbling 9 bits out of 10, at least. And that is with the old disk technology that was out of date before the paper was even written - with newer disks, the recovery rate would be far lower.

Not good enough. Scraping all the magnetic coating off is though.

We used to have a clock ornament made of a platter of a VAX disk that was in a serious head crash with "back up your data" written underneath. The gouges in its surface were quite impressive.

Exactly. You have no way of knowing which bits might belong to which write, or which path was taken by the head.

In some cases, you can build up useful information from just a few bits here and there. But it is critical that you know which bits you know, and you know where they go, and that you can re-use the information in repeated cases. None of that is true in this case.

It is also true that you can often put together some useful information even if there are missing bits - things like ASCII text files have a fair amount of redundancy, and you have a reasonable chance of filling in some gaps or getting a partial result.

But recovery of hard disk data that has been erased is not like that. You can guess a few bits here and there, but are usually not sure if they are valid. You don't know /when/ that old data was written, and can't match it with other old data. At most, you get a rough idea of a small proportion of the bits - not enough to fit together a bigger picture.

A few seconds on a band saw should work fine.

Angle grinder.

Harddisks are surprisingly robust.

See this video

Wim

Not on MY bandsaw. Not in my kitchen oven, either.

At the focus of a solar mirror array, it should be possible to load a few dozen drives an hour, collecting dribbles of aluminum/zinc alloy all the while.

iconductor era

logic circuits

you could presumably construct an identical bismuth-centred arrangement of iron atoms.

equal numbers of bismuth and iron atoms.

These crystal structure images can be hard to appreciate in light of the ch emical formula. The single iron in the center is easy enough to see. The eight bismuth atoms are each only one eighth inside this cube while being s hared with seven other cubes, so this cube only has 1 bismuth atom. The six oxygen atoms are each on the face of the cube, so half-in resulting in onl y 3 oxygen atoms inside the cube giving the formula shown.

Rick C.

Tesla referral code -

A 1" cut on the band saw, just a bit into the platter, should be fine.

Or drill a hole thereabouts.

As a practical matter, I could just mash the pins on the connector. I don't expect the Russian intelligence agencies to get my old drives.

They'd probably be mining more fertile ground.

It's actually the Chinese who are more active in technology espionage - or at least that's what the Australian government always tells us when yet another defense technology firm gets hacked.

On Dec 9, 2018, John Larkin wrote (in article):

The platter may well be made of glass.

My go-to method involves putting the hard drive in a ziplock bag (to contain the fragments), a large boulder (or an anvil), and a 2-pound drilling hammer. This would give even an intelligence agency some trouble.

If you really want to erase the databeyond any possible recovery, raise the temperature above the Curie Temperature of the magnetic alloy on the platters:. Beware the toxic smoke.

Joe Gwinn

My Makita hammer drill would work fine. It chews right into concrete.

Three of the walls of my office are concrete, so I need the Makita to hang or bolt anything.

... fine for WHAT? If one has electron microscope capability, a cut would take out one or two sectora from each full disk track, leaving the remaining data intact. Write-with-randoms and a couple of passes would be much more data-destructive, and would leave you with a working drive.

Bandsaw teeth can be damaged (bent/broken) if they cut into something that subsequently moves... or into a ceramic magnet.