Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 21 Aug 2012 23:23:16 +0400
From: Solar Designer <>
Subject: Re: Cmiyc 2012 : samu writeup

On Mon, Aug 06, 2012 at 10:17:48AM +0200, Samuele Giovanni Tonon wrote:
> * 5 server with 2 Intel(R) Xeon(TM) CPU 3.00GHz for a total of 8 cores
>   per server each and with glusterfs sharing the working dir

As we figured out later, these were Pentium D'ish (probably Xeon 5050).
So that's 2 P4'ish cores per chip, 4 per server, or a total of 20 P4'ish
cores (40 logical CPUs).

These are relatively slow at SSE2 (about twice slower than Core 2 and
newer), but are OK at non-SIMD code (like bcrypt).  For SSE2 code,
those 5 servers combined are roughly equivalent to 10 Core 2'ish cores.

Lesson for next year: find out what hardware each team member is going
to use before we start, have others on the team give recommendations on
optimal use of that hardware (although such recommendations will need to
be revised once the actual hash types and weights are known).  We did
this to some extent this year, but we could do better.


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.