Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Mon, 22 Apr 2013 16:57:34 -0500
From: "jfoug" <jfoug@....net>
To: <john-dev@...ts.openwall.com>
Subject: RE: formats that duplicate dynamics

here are some quick/dirty timings.

$ ../run/john -test=5 -form=postgres
Benchmarking: PostgreSQL MD5 challenge-response [32/32]... (8xOMP) DONE
Raw:    4860K c/s real, 1548K c/s virtual

$ OMP_NUM_THREADS=1 ../run/john -test=5 -form=dynamic_1015
Benchmarking: dynamic_1015 md5(md5($p.$u).$s) [128/128 SSE2 intrinsics
480x4x3]... DONE
Many salts:     8669K c/s real, 8696K c/s virtual
Only one salt:  7435K c/s real, 7432K c/s virtual

Benchmarking: dynamic_1015 md5(md5($p.$u).$s) [128/128 SSE2 intrinsics
480x4x3]... (4xOMP) DONE
Many salts:     26749K c/s real, 7654K c/s virtual
Only one salt:  16951K c/s real, 6915K c/s virtual

So even non-OMP 'unstable', 1015 should much faster than postgres format, on
any reasonable build, simply due to SSE2

Jim.

From: magnum Sent: Monday, April 22, 2013 15:12
>A caveat is that postgres supports OMP while dynamic does not, in unstable.
>
>Bleeding's dynamic does support OMP now though (thanks, Jim!), so this
alternative will become increasingly attractive as this support matures. In
bleeding I think we should start to get rid of thick formats that can be
done with dynamic (and/or a thin format) unless they are significantly
faster than dynamic.
>
>magnum

Powered by blists - more mailing lists

Your e-mail address:

Powered by Openwall GNU/*/Linux - Powered by OpenVZ