Openwall GNU/*/Linux - a small security-enhanced Linux distro for servers
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Sat, 14 Jul 2012 15:13:49 +0400
From: Solar Designer <solar@...nwall.com>
To: Tavis Ormandy <taviso@...xchg8b.com>
Cc: john-dev@...ts.openwall.com
Subject: Re: Rotate and bitselect investigation

On Sat, Jul 14, 2012 at 12:55:47PM +0200, Tavis Ormandy wrote:
> I meant the code in general, the results on amd are dissapointing
> compared to the results on intel.

You probably mean when comparing against AMD CPUs without XOP?

I am not seeing better results on Intel CPUs than what I get on AMD with
XOP.  For example, here's what I am getting on "Intel(R) Xeon(R) CPU
E31230 @ 3.20GHz" (Sandy Bridge, but without AVX), apparently at 3.3 GHz,
gcc 4.6.2:

Benchmarking: Raw SHA-1 (pwlen <= 15) [128/128 SSE2 intrinsics 4x]...  DONE
Raw:    17892K c/s real, 17892K c/s virtual

whereas on FX-8120 the speed is almost 29M c/s (with XOP).  (All of
these are for 1 core, indeed.)

...or are you not talking about your SHA-1 code, but in general?

Alexander

Powered by blists - more mailing lists

Your e-mail address:

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