[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CABob6iowUcw-y2ArU8uKmvd9xVC=uGg3b=oOFgHxkUkhZgyn3Q@mail.gmail.com>
Date: Tue, 18 Jun 2013 03:20:47 +0200
From: Lukas Odzioba <lukas.odzioba@...il.com>
To: john-dev@...ts.openwall.com
Subject: Re: [GSoC] Low-level GPU programming
2013/6/11 Solar Designer <solar@...nwall.com>:
> For JtR, the objective is to speedup some formats, for which we believe
> we're hit by OpenCL limitations the most. These are des-opencl (need
> runtime code patching), maybe bf-opencl, maybe SHA-512 stuff.
Since Sayantan will be working on descrypt and bcrypt I think that
this project should focus on sha512*.
There are several reasons for that:
1) It might be hard to distinguish who did what.
2) Sayantan is more familiar with those formats and there is a lot
chance that Daniel will be forced to duplicate his work.
3) Bad communication could easily waste a lot of time.
By sha512* we mean at least three formats, and first two are the most
revelant to us:
pbkdf2-sha512
sha512crypt
raw-sha512
Before midterm evaluation it would be good to see performance
improvement using low level developement on at least one of those
formats.
To do so you Daniel will have to create some tools for low level
programming, it possibly not be a fully functional assembler but
something what will place foundations to it.
Since Daniel will be absent in week 6, and midterm evaluation is in
week 7 we should achieve that on week 5.
Lukas
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.
