Data recovery from NVMe SSD drives

[oliver1996]

Dear HDDGuru Community,

we are getting still more and more M.2 NVMe SSD drives to data recovery. Many of them we are able to recover with PC3000+DS, but unfortunately more complex cases with (probably) damaged controller are still unsolved by us. What is your experience with these drives and do you have any tips how to get the data back in some more severe hardware cases? Do you swap the controller chip or reball the NAND chips for some cases? Were you successful?

I will be grateful for any tips or advice regarding these drives!

Oliver

[DRUG]

This is a complex topic and usually if you need to ask you won't be able to do it.

For unsupported drives you are on your own.

Some of us spend weeks or months on a specific device and then offer the solution in-house.

You have to understand that most nvme drives have controller tied to nands, so moving nands to a donor is out of the equation.

Also understand:
Nvme CPUs have diferent operation modes.
Try to learn how to manipulate such modes.
Read about voltage injection and jtag.

Every case is diferent, but I wish you luck.

[Arch Stanton]

I don't know if it was shared here, but this video is thought provoking and touches the subject and somewhat underlines what DRUG is saying: https://youtu.be/fuO3t89lQno

[DRUG]

I don't know if it was shared here, but this video is thought provoking and touches the subject and somewhat underlines what DRUG is saying: https://youtu.be/fuO3t89lQno

We understand he as skin in the game and this proves that his tools are a good solution.
He may be biased but the reality is that he is correct in most of the things he says.

Going forward I think big players will offer big $ to engineers involved in manufacturing the most mainstream devices to learn the go arounds and then a public solution will appear.

[DRUG]

I also believe in the possibility that we see ACE et al to start offering solutions in house, much like Cellebrite has CAS, ACE may open an advanced service for it's users.

Big $ are there to be made. Just this week we got two devices that both customers would pay a five digit sum to have the data back. How they can pay this and yet have no backups is out of my understanding

[Lardman]

I also believe in the possibility that we see ACE et al to start offering solutions in house, much like Cellebrite has CAS, ACE may open an advanced service for it's users.

I think you may well have a point there, I got the distinct impression from a recent webcast that they are sitting on solutions for controllers, either because they can't make them generic or they're more involved than a simple software hack.

[DRUG]

I also believe in the possibility that we see ACE et al to start offering solutions in house, much like Cellebrite has CAS, ACE may open an advanced service for it's users.

I think you may well have a point there, I got the distinct impression from a recent webcast that they are sitting on solutions for controllers, either because they can't make them generic or they're more involved than a simple software hack.

Yes, exactly.

For example, Marvell 88SS1074 can be solved, but it is not a clickable solution.

Honestly, the moto stands true: If it's ARM, it can be hacked.