Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Sun, 18 Mar 2012 09:02:16 +0400
From: Solar Designer <solar@...nwall.com>
To: john-dev@...ts.openwall.com
Subject: Re: SSH thread-safety

On Sun, Mar 18, 2012 at 08:39:24AM +0400, Solar Designer wrote:
> BTW, FX-8150 is documented to use 3.6 base / 3.9 full load turbo / 4.2
> half load turbo, so maybe I should try 3.9 / 4.2.

Seems to work at 3.9 / 4.5.  3.9 matches FX-8150's, and the 600 MHz
difference is consistent with FX-8120's default (was 3.4 to 4.0).
So both frequencies are now 500 MHz higher than rated for this CPU.
I haven't really tried going beyond 4.5 yet (except for the experiment
up to 5.1 GHz in BIOS screens only), but the full load frequency is more
important for JtR anyway (since we're normally going to use all cores),
and 3.9 GHz appears to be the max for this chip without tweaking other
settings (voltage, base clock).

Non-OpenMP:

Benchmarking: ssh [32/64]... DONE
Raw:    164611 c/s real, 164611 c/s virtual

OpenMP with 8 threads:

Benchmarking: ssh [32/64]... (8xOMP) DONE
Raw:    795785 c/s real, 99567 c/s virtual

795785/(164611*8*3.9/4.5) = 70% efficiency (of true 8-core CPU and
zero-overhead parallelization)

Alexander

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.