Nice. Still the problem with more current drives is the extreme data density in comparison to previous generations and thus the low tolerances during the process. In practise you would probably need to scan all the whole platters anyways as you don't really know where the file is and also the raw data might have been written over several ones in paralell to increase read/write speed (or so I assume).
True, and you're not going to get the data off with a simple Rpi, but the concept should still work. The company I mentioned recovers data from crashed drives all the time for $1k or so, they'd be able to give it a good go.
Quote
The reason I think the Ironkey guy has a better chance is mostly because a) He already has the device b) It is not damaged and c) the guy (presumably) already has enough money to fund the costly effort.
As for the technique, I was thinking about maybe using a SEM to recover either the "access password" or the internal encryption key + separately dump the flash memory (which I assume is a regular one). Something like this:
https://www.cl.cam.ac.uk/~sps32/cardis2016_sem.pdf
Possible. Some of these devices do have anti-tamper that will shatter the chip if you try to pry it or stuff like that. Still we're taking a budget of ~100 million to crack it which is well within the above paper's recommendation of "just buy a fucking ion mill and 3d the state of the chip". The fact he hasn't done something like this makes me wonder if it's all a bullshit story.As for the technique, I was thinking about maybe using a SEM to recover either the "access password" or the internal encryption key + separately dump the flash memory (which I assume is a regular one). Something like this:
https://www.cl.cam.ac.uk/~sps32/cardis2016_sem.pdf
Quote
P.S.: Or maybe just dump the contents of the main flash chip and try to brute force the encryption key IF it is derived from the "access password".
I'm guessing the anti-tamper options will limit this. Then again even if the device is set back to zero there might be residual info to reconstruct the data. This is why the military grade ones literally explode. We are talking a budget of 100m to break it, and every safe is rated not to keep your data secure forever but as a function of "time and tools". He should at least be able to get to the "copy the data and brute the key" level with current tech.