0
$\begingroup$

I have been working on an assignment to solve argon2i hashes but unfortunately. I am not able to use GPU with JTR i.e. no opencl mode for argon2. Thus the speed is very slow. Any suggestion how to proceed using GPU based attack?

$\endgroup$
0
$\begingroup$

I am not able to use GPU with JTR i.e. no opencl mode for argon2. Thus the speed is very slow. Any suggestion how to proceed using GPU based attack?

This is exactly one of the reasons that argon2 exists.

It was explicitly designed so that hardware other than a general purpose CPU provides only a minimal advantage, if any.

Proceeding with GPU based attacks is probably a waste of a GPU. One of the core features of argon2 is a large amount of memory usage, which GPUs are not exceptionally good at.

Options

Your best bet is to have a large number of normal computers with sufficient RAM, or investigate FPGAs. FPGAs may not be able to provide a worthwhile advantage due to the memory requirement, and they will not be so easy to work with (unless you are already skilled with an HDL).

Neither option will be cheap, and I assume that they are not practical in your situation.

Most likely, you are stuck testing one password guess at a time using your CPU. In such a case, the best that you can do is to use a good word list and hope that the password is weak. Either that, or find some alternative method of finding the password that doesn't involve cracking argon2 hashes.

$\endgroup$
  • $\begingroup$ Thank you so much for the quick response. :) I almost hit a threshold for trying. I think I understand you now. I hope this thread would be helpful for new noobs like me. Just one more question, does scaling cpu makes a huge difference or not? Example using a high end i7 rather then lower end dual core? $\endgroup$ – hashcracker101 Oct 11 '18 at 23:01
  • $\begingroup$ @hashcracker101 I'm not sure, it may be limited by memory access times rather than CPU clock speed. You'd probably have to test it to be sure. I would not be surprised if the i7 was faster, but I would be surprised if it was faster by a huge amount (e.g. I wouldn't expect it to be 10x as fast). But I can't really say. $\endgroup$ – Ella Rose Oct 11 '18 at 23:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.