[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Sat, 27 Jul 2013 19:11:25 +0400
From: Solar Designer <solar@...nwall.com>
To: john-dev@...ts.openwall.com
Subject: Re: Sayantan's weekly report #6
On Sat, Jul 27, 2013 at 06:18:33PM +0530, Sayantan Datta wrote:
> On Sat, Jul 27, 2013 at 2:44 PM, Lukas Odzioba <lukas.odzioba@...il.com>wrote:
> > Last week solar suggested adding NTLM, MSCash (DCC), LM. I would add
> > sha256/512 to this list with lower priority.
> >
> > Did you use TS to test your code?
> > What is the status of merging mask-mode with bleeding?
>
> Isn't nt-opencl same as NTLM ? I'm working on dcc right now. Also LM would
> require a lot of work so I'm lowering its priority, after the contest.
> sha256/512 are on my list.
Note that there's relatively little speedup to be had for SHA-512 by
having mask mode on GPU - maybe "only" 2x or so, compared to having
those candidate passwords fed by the CPU. So I think SHA-512's mask
mode on GPU is not a priority to be done before the contest. Out of the
SHA-*'s, SHA-1 is more important (most speedup possible), next is
SHA-256, and only then SHA-512 and Keccak (slower, so less speedup).
There's also raw MD4 (huge speedup, but relatively rare hash - yet may
be seen in the contest). And I hope we won't see SHA-0. ;-)
OK to postpone LM. I thought it would not be that difficult for you due
to your experience with descrypt, but overall I agree it's less trivial
(and, besides, it's quite fast on CPU).
Alexander
Powered by blists - more mailing lists
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
