encryption with shreddeding function.

[birchgauze]

I read the other day, that encryption of hardware, is becoming increasingly useless do to the computational power of GPUs.

 

So to increase my security I was thinking if there is any encryption software that "breaks" or just starts to shreds once files if there has been to many attempts of breaching the system?

[WereCatf]

6 minutes ago, birchgauze said:

So to increase my security I was thinking if there is any encryption software that "breaks" or just starts to shreds once files if there has been to many attempts of breaching the system?

That'd be useless. When someone is cracking the encryption of something, they're not using the original software, so the original software having such an option wouldn't matter. Encryption is the algorithm, not the software that makes use of the algorithm and algorithms are just maths -- there is no counter for how many times one has failed to decrypt something.

[mariushm]

Someone would clone the drive or make a backup of the files and then run everything in sandbox / virtual machine, and run multiple instances in parallel / reboot as needed. 

Also, a well motivated person could prevent the software from destroying anything, for example the easiest thing that comes to mind is making a custom sata controller in a fpga so that all the commands the software sends to the sata drive get filtered ... so any write/erase command would be ignored.

 

 

[birchgauze]

Arh make sense.

 

[trag1c]

As @WereCatf and @mariushm have said its useless.  If you're truly concerned about brute force attacks then use an algorithm that uses more bits for the key length so that it takes that much longer because at the end of the day the only mitigation to brute force is time. You can't protect against it so all you can do is slow em down.

[WereCatf]

3 minutes ago, trag1c said:

If you're truly concerned about brute force attacks then use an algorithm that uses more bits for the key length so that it takes that much longer because at the end of the day the only mitigation to brute force is time.

If one were to go that far, they should choose an algorithm that's also resistant against quantum crypto - attacks as well, like e.g. one of the solutions mentioned at https://en.wikipedia.org/wiki/Post-quantum_cryptography#Comparison

[birchgauze]

8 minutes ago, WereCatf said:

If one were to go that far, they should choose an algorithm that's also resistant against quantum crypto - attacks as well, like e.g. one of the solutions mentioned at https://en.wikipedia.org/wiki/Post-quantum_cryptography#Comparison

Very interesting, thank you.