Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Mon, 17 Jun 2013 06:35:25 +0400
From: Solar Designer <solar@...nwall.com>
To: john-dev@...ts.openwall.com
Subject: Re: PBKDF2-SHA512

Lukas -

On Mon, Jun 17, 2013 at 03:56:44AM +0200, Lukas Odzioba wrote:
> I started porting my pbkdf2-sha512 cuda code to opencl.

Sounds good.  Thanks!

> It looks like this code running on cpu without touching SSE is 14%
> faster than current cpu version .
> 
> john -test -format=PBKDF2-HMAC-SHA512
> Benchmarking: PBKDF2-HMAC-SHA512, GRUB2 / OS X 10.8 [PBKDF2-SHA512
> 128/128 SSE4.1 2x]... DONE
> Raw:	92.6 c/s real, 92.6 c/s virtual
> 
> john -test -format=grub
> Benchmarking: grub, grub [PBKDF2-SHA512]... DONE
> Raw:	106 c/s real, 106 c/s virtual
> 
> Both tested on the same test vector and cpu.

I think you're comparing one core with C+intrinsics vs. all cores with
OpenCL.  Try building with OpenMP enabled for a more relevant comparison.

Alexander

Powered by blists - more mailing lists

Your e-mail address:

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