Crptohaze to substitute openbsd's implementation of MD5

Discussion and support for the CUDA Multiforcers (Windows and Linux)
  • Ads

Crptohaze to substitute openbsd's implementation of MD5

Postby frede.sch » Fri Oct 12, 2012 12:49 am

Hey,

I'm implementing a hash table of MD5 in C and I'm using the openbsd implementation of MD5 algorithm. I would like to use your fantastic cryptohaze instead of it (I have a core I7 and a GTX 675M and openbsd's implementation is not using any of them appropriately).

Right now I'm generating numbers and hashing them, one after other until the end is reached. Is there an easy way to include cryptohaze and make it generate the hashes? E.x: I give the beginning and the end and the charset of what I want to be hashed and it gives me back the hashes. Or I give it an array of what I want to be hashed and a second one to it stores the hashes.
frede.sch
 
Posts: 7
Joined: Fri Jun 01, 2012 12:45 am

Re: Crptohaze to substitute openbsd's implementation of MD5

Postby Bitweasil » Fri Oct 12, 2012 1:56 am

That's not a currently supported implementation, but it certainly could be modified to do it.

How much data are you hashing?
Bitweasil
Site Admin
 
Posts: 912
Joined: Tue Jan 20, 2009 4:26 pm

Re: Crptohaze to substitute openbsd's implementation of MD5

Postby frede.sch » Fri Oct 12, 2012 2:10 am

Initially I'm hashing until 9 digits numeric, but I'm planning to hash any charset with less possibilities than 2 ^ 64.
frede.sch
 
Posts: 7
Joined: Fri Jun 01, 2012 12:45 am

Re: Crptohaze to substitute openbsd's implementation of MD5

Postby Bitweasil » Fri Oct 12, 2012 2:38 am

Where are you going to store this data? And do you have any idea how much data you're talking about?

And have you considered rainbow tables? I have MD5 rainbow tables available for full US charset through length 8...

Hashing numeric-9 is 1000000000 possibilities.

At 16 bytes of hash + 9 bytes of plain, you're at 1000000000 * (16 + 9) = 25000000000 bytes, or 23GB.

Obviously it can be stored more efficiently, but if you're going to store 2^64 hashes, you're looking at storage beyond the exabyte range.

What, specifically, is your goal? If it's being able to crack unsalted hashes, rainbow tables (which I have an implementation of) are a great solution, and if you have the resources to throw at it, I'm happy to work on my len9 tables.
Bitweasil
Site Admin
 
Posts: 912
Joined: Tue Jan 20, 2009 4:26 pm

Re: Crptohaze to substitute openbsd's implementation of MD5

Postby frede.sch » Fri Oct 12, 2012 3:06 am

Yes, I do have an idea of how much it is.

I'm working on an alternative for rainbow tables, I'm creating a hash table that is more efficient than rainbow tables.
frede.sch
 
Posts: 7
Joined: Fri Jun 01, 2012 12:45 am

Re: Crptohaze to substitute openbsd's implementation of MD5

Postby Bitweasil » Fri Oct 12, 2012 3:52 am

Are you willing to discuss it over email? If it's a solid idea, I'd be happy to help implement it.

The Multiforcer isn't really the framework I'd build it on, though. The GRT tool chain is much better suited to this type of work, and could be trivially modified to do something with hash tables.
Bitweasil
Site Admin
 
Posts: 912
Joined: Tue Jan 20, 2009 4:26 pm

Re: Crptohaze to substitute openbsd's implementation of MD5

Postby frede.sch » Fri Oct 12, 2012 3:58 am

just mailed you on bitweasil at gmail dot com
frede.sch
 
Posts: 7
Joined: Fri Jun 01, 2012 12:45 am


Return to CUDA Multiforcers

Who is online

Users browsing this forum: No registered users and 1 guest

cron