[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CA+E3k92QMfffp=F5405AZ5Ns6z+i=4qfiC+T7QXARx8TsqRecA@mail.gmail.com>
Date: Wed, 6 Jul 2016 18:02:13 -0800
From: Royce Williams <royce@...ho.org>
To: john-users@...ts.openwall.com
Subject: Re: best way to attack bcrypt, for me
On Wed, Jul 6, 2016 at 3:46 PM, Rich Rumble <richrumble@...il.com> wrote:
> I am aware of the great work Openwall has done with FPGA's and other
> form factors when it comes to bcrypt (I've watched the video, seen the
> slides). Let's say I'm an average user of computers, no programming
> skills, I can run ./configure && make clean && make, 90% of the time
> it will work :)
>
> I want to get as many Bcrypts as possible in the shortest amount of
> time (naturally) but I might not be able to setup and use a Zynq. If
> you take "ease of use" into consideration, which platform offers the
> best result for the money.
> I might be able to do 20k c/s with a Zynq 7045, but will "I" be able
> to set that up?
>
> I'm budgeting $2,000 (ish). That is 2 maybe 3 good GPU's, or one Zynq
> board, I could buy 2-3 Phi units, a handful of others as well.
>
> 2k + ease of use (aka stupid user) = ....?
>
> I've not used a MIC or FPGA, or GPU for that matter :)
> Again it's not necessarily efficiency I'm after, it doesn't have to be
> the best setup for the $$, I might technically get more with a Zynq
> for instance, but if a Haswell if two i7's will do a decent job and be
> easier to use, I'd probably go with i7 (for example :)
> -rich
ZTEX 1.15y clones would be cost-effective for bcrypt, both for initial
investment and (obviously) for power consumption. Do they have
proof-of-concept status with JtR yet? If so, how much effort would it
take to make them configure/make-capable?
Royce
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.
